COPYRIGHT TRIBUNAL OF AUSTRALIA

Phonographic Performance Company of Australia Limited under s 154 of the Copyright Act 1968 (Cth) [2015] ACopyT 3

THE TRIBUNAL DIRECTS THAT:

1.    The parties confer and file, within 21 days, agreed or competing directions for the further conduct of the matter.

2.    The matter be listed for directions on a date to be determined in consultation with the parties within 14 days thereafter.

3.    Pending further or other order, neither party is to file any document (including any affidavit) before the directions hearing, without leave of the Tribunal.

TABLE OF CONTENTS

REASONS FOR DETERMINATION

1.    THE TRIBUNAL’S TASK

1.1    Some agreed background matters

1    The Tribunal’s task is vested in it by ss 154(1) and (4) of the Copyright Act 1968 (Cth) (the Copyright Act) which provide that:

(1)    Where a licensor proposes to bring a licence scheme into operation, he or she may refer the scheme to the Tribunal.

…

(4)    The Tribunal shall consider a scheme referred under this section and, after giving to the parties to the reference an opportunity of presenting their cases, shall make such order, confirming or varying the scheme or substituting for the scheme another scheme proposed by one of the parties, as the Tribunal considers reasonable in the circumstances.

2    Phonographic Performance Company of Australia Limited (PPCA), a collecting society representing the interests of record companies and recording artists, is a licensor as defined in s 136(1) of the Copyright Act. PPCA referred a scheme to the Tribunal under s 154(1). Pursuant to that scheme PPCA, as the holder of a non-exclusive licence in respect of sound recordings from record companies and recording artists, would grant to commercial radio broadcasters a licence to use sound recordings in simulcasts. A simulcast involves the transmission of the radio broadcast, including the sound recordings used in that broadcast, to listeners using the internet.

3    Commercial Radio Australia Limited (CRA), an industry body representing the interests of around 259 commercial radio stations in both metropolitan and regional areas, became a party to the reference as contemplated by ss 154(2)(b) and (3) of the Copyright Act which enable the Tribunal to make an organisation or person with a substantial interest in the operation of the scheme a party to the reference. CRA proposed different schemes from that of PPCA relating to the licensing of the same rights, namely, to enable commercial radio broadcasters to use sound recordings in the simulcast of their broadcasts.

4    The referral of these schemes reflects the decision in Phonographic Performance Company of Australia Ltd v Commercial Radio Australia Ltd [2013] FCAFC 11; (2013) 209 FCR 331 in which the Full Court of the Federal Court of Australia decided that simulcasting was not within the scope of an existing agreement between PPCA and CRA under the Broadcasting Services Act 1992 (Cth) (the Broadcasting Services Act). As the Full Court explained, referring to the agreement between PPCA and CRA as the Industry Agreement (but mostly identified in the reasons below as the PPCA-CRA Broadcast Agreement):

[2] The Industry Agreement relates to the grant, by PPCA to CRA members, of licences in relation to certain sound recordings (the Sound Recordings). PPCA is the collecting society in respect of the Sound Recordings for the purposes of the Copyright Act 1968 (Cth) (the Copyright Act). Members of CRA are the holders of commercial radio broadcasting licences granted under the Broadcasting Act. Under the Industry Agreement, the licences are granted in accordance with, and subject to the terms of, the form of an agreement set out in a schedule to the Industry Agreement (the Member Agreement).

[3] Under the Member Agreement, PPCA grants to the CRA member a non-exclusive licence of the Broadcasting Right, as defined in the Member Agreement. The Broadcasting Right is defined in the Member Agreement as the right, from time to time, to broadcast the Sound Recordings in Australia…

[4] Under s 6(1) of the Broadcasting Act, broadcasting service means, relevantly, a service that delivers radio programs to persons having equipment appropriate for receiving that service, but does not include, relevantly, a service that the Minister determines not to fall within the definition. On 12 September 2000, the Minister made a determination under that provision (the Ministerial Determination) that a class of services described in the Ministerial Determination does not fall within the definition.

[5] The effect of the definition, as affected by the Ministerial Determination, is that broadcasting service means, relevantly, a service that delivers radio programs to persons having the equipment appropriate for receiving that service, where the delivery uses the radiofrequency spectrum, cable, optical fibre, satellite or any other means or a combination of those means, but does not include a service that makes available radio programs using the internet, other than a service that delivers radio programs using the broadcasting services bands…

[6] From 2001, at the latest, some CRA members have, simultaneously with transmission using the broadcasting services bands, streamed radio programs on the internet that include the Sound Recordings. The content of the radio program made available using the internet is identical to the content of the radio program delivered using the broadcasting services bands…The question in the appeal is whether the making available of radio programs using the internet in those circumstances is within the licence granted by PPCA under the Member Agreement. PPCA contends that it is not. CRA, on behalf of its members, contends that it is.

5    The Full Court resolved this dispute at [66] by concluding that:

A broadcasting service is not a service that makes radio programs available using the internet (due to the exclusion in the Ministerial Determination), unless it is a service that delivers radio programs using the broadcasting services bands (due to the exception to the exclusion in the Ministerial Determination): thus, only some services that make radio programs available using the internet are excepted from the exclusion in the definition, namely, services that deliver radio programs using that part of the radiofrequency spectrum that consists of the broadcasting services bands.

6    There is no dispute between the parties that the schemes referred to the Tribunal in the present case relate to the use of the sound recordings forming part of commercial radio broadcasts by simulcast of the radio program over the internet, being a use which does not involve the broadcasting services bands. In other words, such a use does not form part of the broadcasting service licensed pursuant to the PPCA-CRA Broadcast Agreement.

7    We record below some further uncontroversial background material which sets the context for the Tribunal’s task. This background material is taken from the written submissions of the parties, with some adaptations where appropriate.

8    Radio stations commenced operating in Australia in the 1920s and 1930s. Their primary activity is the delivery of radio programs to the public. The Australian Communications and Media Authority (ACMA) regulates the grant of and compliance with radio broadcast licences.

9    The 259 radio stations which CRA represents consist of approximately:

(a)    39 in metropolitan areas (major cities);

(b)    12 in regional (smaller) cities; and

(c)    208 in country areas,

of which, there are approximately 157 FM radio stations and 102 AM stations.

10    Radio stations use music in their programs to greater or lesser extents depending on the type of program.

11    For copyright purposes, each piece of music is comprised of a sound recording (the use of which is licensed to the radio station by PPCA) and an underlying musical work (licensed by APRA/AMCOS, the Australasian Performing Rights Association and the Australasian Mechanical Copyright Owners Society).

12    A radio program is an audio stream. Before digital technology the only way for a radio program to be distributed was by transmission over the radio frequency spectrum (that is, by broadcast) to FM transmitters, AM transmitters and DAB+ transmitters. Digital technology enables the radio audio stream to be distributed over the internet. From 2001 many radio stations began to use the internet to deliver to listeners by that means the radio program being broadcast. This is radio simulcasting. Simulcasts of radio programs may be received wherever the internet is available – either through a “listen live” link, often located on a station’s website accessed on a computer, or a mobile app accessed on a smartphone or tablet.

13    The basic business model of radio stations is that the programs are provided free of charge to the audience, the radio stations being “ad funded”. Advertisers pay the stations to insert advertisements into their radio programs in order to reach their desired audience. Generally speaking, at present, advertising is priced and sold based on audience size and reach. Radio stations in Australia do not currently offer different advertising on broadcasts and simulcasts to advertisers. Advertisements in radio programs account for between about 95% and 97% of Australian radio stations’ revenue at present.

14    Radio stations do not charge listeners for receiving the radio program via either the broadcast or the internet. The simulcast of radio is not subscription based. Radio is freely available to listeners including by simulcast if they have internet access. This contrasts with the circumstances before the Tribunal in two other matters discussed below (referred to as the Nightclubs case and the Gyms case).

15    For metropolitan and large regional stations ratings (which determine the audience share of each station) are a key determinant of what advertisers are willing to pay for advertising time. Radio Ratings Surveys are conducted by GfK Retail and Technology Australia Pty Ltd (GfK) eight times per year in large metropolitan centres and three times a year in other metropolitan/large regional centres.

16    Some 156 radio stations were simulcasting before this matter commenced. Many regional radio stations, however, ceased simulcasting when the interim scheme was made. The interim scheme is a scheme adopted by the Tribunal on 16 December 2013 for the period pending the resolution of this matter. Subsequently, a few regional radio stations re-commenced simulcasting.

17    The appropriate approach of the Tribunal to its task was not in material dispute.

18    As CRA submitted, there is no presumption in favour of PPCA’s scheme because it is a licensor and referred its scheme first so as to enliven the Tribunal’s jurisdiction. As CRA put it, and we accept:

[the] schemes must be considered on their merits according to the statutory test of reasonableness in the circumstances. The preference of copyright owners is no more relevant than the preference of licensees. While there is no Australian authority on point, the Copyright Tribunal in the United Kingdom has confirmed that no such presumption exists and the principle has been applied in New Zealand where the corresponding provision in the Copyright Act is in similar terms to s.154 [British Phonographic Industry Limited and Ors v Mechanical Copyright Protection Society Limited and Ors, CT84-90/05, 19 July 2007 at [44]-[46], especially [46]; British Phonographic Industry Limited v Mechanical Copyright Protection Society Limited (no 2) [1993] EMLR 86 applied in Phonographic Performances (N.Z.) Ltd v Radioworks Ltd and anor [2010] NZCOP 1 at [42]].

19    As CRA also observed, the history of previous negotiations and dealings is an important consideration in determining what is reasonable in the circumstances (Reference by Australasian Performing Right Association Ltd; Re Australian Broadcasting Corporation (1985) 5 IPR 449 at 466).

20    Further, as PPCA submitted, there has not been any authoritative exposition of the meaning of “reasonable in the circumstances” by the Federal Court. Despite this, previous decisions of the Tribunal are instructive. In Reference by Phonographic Performance Company of Australia Limited under s 154(1) of the Copyright Act 1968 [2007] ACopyT 1; (2007) 73 IPR 162 (the Nightclubs case), the Full Tribunal said:

[10] In each reference under s 154, the tribunal must make a value judgment as to what it considers reasonable in the circumstances. It is not usually possible to calculate mathematically the correct licence fee in any particular case: Australasian Performing Right Association Ltd v Federation of Australian Radio Broadcasters Ltd (1999) 46 IPR 20; [1999] ACopyT 4 at [11] and [19] (APRA v AFRB). Where approval of a scheme would lead to a substantial increase in fees, the increases can be phased in over a period of years rather than being introduced immediately. In the present case, the society proposes that there be a phasing in of the increase claimed by it.

[11] In determining whether a proposed scheme, and the licence fee payable under it, are reasonable, a number of approaches might be adopted. The approaches include the following, which may overlap to a certain extent:

•    Market rate: the rate actually being charged for the same licence in the same market in similar circumstances.

•    Notional bargain rate: the rate on which the tribunal considers the parties would agree in a hypothetical negotiation, between a willing but not anxious licensor and a willing but not anxious licensee.

•    Comparable bargains: bargains not in the same market but sufficiently similar to such a notional bargain as to provide guidance to the tribunal.

•    Judicial estimation: the rate determined by the tribunal after taking into account a range of matters such as:

•    previous agreements or negotiations between the parties;

•    comparison with other jurisdictions;

•    comparison with rates set by other licensors, capacity to pay, value of the copyright material, the general public interest and the interests of consumers; and

•    administrative costs of a licensing body: see Audio-Visual Copyright Society Ltd v Foxtel Management Pty Ltd (No 4) (2006) 68 IPR 367 at [131] and [142].

21    The Full Tribunal also noted at [13] that:

Section 157(2) of the Copyright Act allows a person to apply to the tribunal if that person claims that the grant of a licence would be subject to charges or conditions that are not reasonable in the circumstances of the particular case. If the tribunal is satisfied that such a claim is well founded, the tribunal may make an order specifying the charges and conditions that it considers reasonable in the particular circumstances. Thus, the tribunal is empowered to address the unreasonable application of a scheme in a particular case.

22    At [200] the Full Tribunal said this:

If it be the fact that the right to play recorded music has the value claimed by the society, the fact that many nightclubs presently operating cannot afford to pay for that privilege is not a reason for them to be subsidised by those whom the society represents. Ultimately, market forces will operate. That is to say, one of the consequences of the introduction of the proposed licence fee may be to reduce significantly the number of nightclub operators who are prepared to pay the fee for the privilege of playing recorded music at their venues. Inefficient operators who are required to pay a market price for all of the services they require in order to conduct their businesses may be forced out of business. That is the nature of a competitive market.

23    If a scheme involves the provision of a forced subsidy by a licensor to a licensee then it may be accepted that, in the ordinary course, such a scheme fails the test of being reasonable in the circumstances. It is difficult to conceive of circumstances which would make such a forced subsidy reasonable. Equally, however, if no prospective licensee can afford to pay the value claimed by the licensor so that an otherwise apparently useful activity cannot be carried out at all, it tends to suggest that the rate claimed does not reflect a notional bargain into which the parties rationally could have entered.

24    The concept of a notional bargain was considered by the Full Tribunal in WEA Records Pty Ltd v Stereo FM Pty Ltd (1983) 1 IPR 6 (the 2MMM case) in which the issue was the amount that should be payable by a broadcaster (in that case, the FM commercial radio broadcaster, 2MMM) for the broadcast of sound recordings. The Full Tribunal focused on the value to 2MMM of the rights to be licensed, albeit recognising that those rights were confined to what is known as protected sound recordings only. The notional bargain approach, as the Full Tribunal decided, was to be applied to:

a collective bargain by the broadcaster on the one hand and all identifiable copyright owners on the other, each exercising their actual bargaining power. The only assumption which is made is that they are each willing to negotiate and conclude an agreement.

25    In other words, insofar as possible, the notional bargain must reflect the actual position of the parties, not (as in a compulsory acquisition case) the position that would be reached between a hypothetical licensor and a hypothetical licensee with notionally equal bargaining power.

1.2    The significant disagreed matters

26    Virtually every other issue between the parties was in dispute.

27    Indeed, the nature and extent of the disputes in this matter indicate that, but for statutory necessity, the required assumption of a willingness to negotiate and conclude an agreement could not be made.

28    The disputes extended to the most peripheral of issues (such as the extent of FM chips in mobile phones to enable people to listen to radio on that device by the broadcast rather than the simulcast transmission), as well as other issues incapable of rational resolution by mere assertion and counter assertion (such as the extent of benefit to record companies from radio stations playing songs). None of this prevented the parties from adducing swathes of evidence about the insignificant or imponderable. It should be understood that we have no intention of referring to all of the evidence with which we were burdened because we found so much of it unnecessary and unhelpful.

29    Focusing only on the most significant issues, the following matters became apparent through the course of the hearing:

As to the relevant parties to the scheme:

(1)    PPCA would not grant a licence to CRA. PPCA’s scheme involves the grant of a licence by PPCA to each person who holds a commercial radio broadcasting licence under the Broadcasting Services Act.

In this regard, it should be noted that a “licence scheme” is defined in s 136(1) of the Copyright Act in these terms:

licence scheme means a scheme (including anything in the nature of a scheme, whether called a scheme or tariff or called by any other name) formulated by a licensor or licensors and setting out the classes of cases in which the licensor or each of the licensors is willing, or the persons on whose behalf the licensor or each of the licensors acts are willing, to grant licences and the charges (if any) subject to payment of which, and the conditions subject to which, licences would be granted in those classes of cases.

According to PPCA it could not be reasonable to require it to enter into an agreement with a party with which it did not wish to deal, being CRA, when the object of the legislative provisions was to enable the referral of a scheme, thus enlivening the Tribunal’s jurisdiction, in respect of “the classes of cases in which the licensor or each of the licensors is willing, or the persons on whose behalf the licensor or each of the licensors acts are willing, to grant licences”. CRA was not such a person, not just because PPCA was not willing to deal with CRA, but because CRA was not a person to which a licence would be granted in any event.

(2)    CRA contended that the scheme (or schemes, that too being a point in dispute of no ultimate significance) it had referred to the Tribunal, which provided for a structure the same as the PPCA-CRA Broadcast Agreement (that is, a head agreement between PPCA and CRA, with licences granted to members of CRA under that head agreement) or the appointment of CRA as the agent for the licensee for the purpose of payments of fees to PPCA, would be reasonable in the circumstances.

As to the structure of the scheme:

(3)    PPCA contended that it was reasonable in the circumstances that the charge to which commercial radio broadcasters would be subject for their simulcast of relevant sound recordings (being those the subject of PPCA’s power to licence) be calculated on a per stream basis. That is, the commercial radio broadcasters would pay a fixed amount each time such a sound recording was streamed via an internet connection.

(4)    CRA contended that a scheme based on a per stream approach would be unreasonable for numerous reasons including that:

    simulcasting is the transmission of the same radio broadcast over the internet. Radio stations pay a fee to PPCA for the broadcast of the radio program on a percentage of revenue basis and have done so for many years, it not being possible for payment on a per stream basis to be imposed on the broadcast;

    radio stations have in place long-standing arrangements enabling them to pay to PPCA licence fees on a percentage of revenue basis for the broadcast rights, which arrangements will continue irrespective of the simulcast licence. As such, payment for the simulcast licence on a per stream basis would involve imposing a new payment assessment and collection system for an activity (simulcasting) dependent on and ancillary to another activity (broadcasting);

    whatever the future of simulcasting, it is presently not the dominant mode by which listeners listen to the radio, with the percentage of listeners listening via the internet presently being somewhat over 10%;

    the imposition of a new payment assessment and collection system on a per stream basis for simulcasting is not readily able to be implemented by all radio stations and each station, some to a small extent but others to a much greater extent, will have to invest time, money and resources into the development of new systems to enable such a system to operate, with some being incapable of so doing other than by contracting out the function to third party providers which do not currently operate in Australia; and

    the percentage of revenue approach reflects the fact that radio stations make nearly all of their money out of advertising. Listeners do not pay to listen to the radio. Advertisers pay to place advertisements within the radio program. Advertisement spots are priced by reference to audience numbers which are assessed for metropolitan and some regional stations by ratings. These spots are priced by reference to the total audience share of the radio station and not the proportion listening to the broadcast as opposed to the simulcast. Moreover, the percentage of revenue paid under the PPCA-CRA Broadcast Agreement is based on gross revenues, not revenues derived only from the broadcast, there being no such distinction drawn by radio stations and advertisers.

As to the amount of any charge for a per stream scheme:

(5)    PPCA formulated proposed charges based on the current arrangements in the United States (the US), in particular, an agreement between SoundExchange or SX (the US equivalent of PPCA) and the National Association of Broadcasters (NAB) (the US equivalent of CRA) dated March 2009, which applies to all internet audio transmissions by commercial radio broadcasters (including simulcasting) (the NAB-SX Agreement). PPCA’s proposed charges, which were set below the US rates, are as follows:

PPCA also proposed an alternative arrangement for any regional radio station with a licence area of 2% or less of the Australian population by which such a station could pay a flat fee of $5000 per year provided the station capped its number of concurrent users of the station’s simulcast at 100. In any such case, the reporting and audit requirements necessary to be implemented to enable a per stream scheme to operate would not apply to that station.

This approach was supported as reasonable by two economists called by PPCA, Dr Eisenach and Mr Massarsky.

(6)    CRA also called an economist, Dr Epstein, and a valuer, Mr Samuel, who challenged the rates proposed by PPCA as too high and the range of agreements relied on by PPCA as lacking comparability. Ultimately, however, neither expert nor CRA itself proposed any alternative per stream rates. Instead, CRA’s position was that any attempt to determine a reasonable per stream rate is doomed because there is no proper evidentiary foundation to determine such a rate.

As to the amount of any charge for a percentage of revenue scheme:

(7)    Dr Epstein and Mr Samuel proposed calculations of the appropriate percentage of revenue for simulcasting radio programs based on the average percentage of revenue of 0.4% paid by radio stations (via CRA) to PPCA under the PPCA-CRA Broadcast Agreement.

On this basis Dr Epstein’s approach resulted in charges calculated as follows:

Gross Licence Fee means, for a Financial Year:

(a)    0.341% x Simulcast Audience Percentage x Members Gross Revenue; and

(b)    0.741% x Members Simulcast Revenue.

Where:

Members Gross Revenue means the gross earnings during that Financial Year of all commercial radio broadcasters licensed pursuant to the provisions of the Broadcasting Services Act 1992 which were Members during that Financial Year.

Members Simulcast Revenue means any revenue earned during a Financial Year by Members which is directly attributable to the Simulcast (such as pre-roll advertising) and is not already included in Members Gross Revenue.

Simulcast Audience Percentage means the proportion of total listeners to Radio Stations who listen via the Simulcast (as opposed to the Broadcast) in the five capital cities (Sydney, Melbourne, Brisbane, Adelaide and Perth) for Surveys 1-8 in the calendar year immediately preceding each Financial Year, as determined by CRA's official audience measurement provider, which at the commencement of this Agreement is GfK Retail and Technology Australia Pty Ltd.

Mr Samuel’s approach resulted in charges calculated as follows:

Gross Licence Fee means:

(a)    in the 2015/16 Financial Year, 0.047% of Members Gross Revenue; and

(b)    for each subsequent Financial Year, the Applicable Rate applied to Members Gross Revenue.

Where:

Applicable Rate means, for each Financial Year, the rate applied in the previous Financial Year plus or minus 0.0034% for each 1% shift in the Simulcast Audience Percentage.

Members Gross Revenue means the gross earnings during that Financial Year of all commercial radio broadcasters licensed pursuant to the provisions of the Broadcasting Services Act 1992 which were Members during that Financial Year.

Simulcast Audience Percentage means the proportion of total listeners to Radio Stations who listen via the Simulcast (as opposed to the Broadcast) in in the five capital cities (Sydney, Melbourne, Brisbane, Adelaide and Perth) for Surveys 1-8 in the calendar year immediately preceding each Financial Year, as determined by CRA's official audience measurement provider, which at the commencement of this Agreement is GfK Retail and Technology Australia Pty Ltd.

It will not be obvious but these formulae result in a fee range for the simulcast right of between 0.0341% and 0.065% of gross revenue, using the current internet listening audience average of 10.7%.

(8)    PPCA, Dr Eisenach and Mr Massarsky contended that the percentage of revenue rate set in the PPCA-CRA Broadcast Agreement is not only out of date but is also artificially constrained by s 152(8) of the Copyright Act which provides that:

(8)    The Tribunal must not make an order that would require a broadcaster who is:

(a)    the holder of a licence allocated by the Australian Communications and Media Authority under the Broadcasting Services Act 1992 that authorises the holder to broadcast radio programs; or

(b)    a person authorised by a class licence determined by that Authority under that Act to broadcast radio programs;

to pay, in respect of the broadcasting of published sound recordings during the period covered by the order, an amount exceeding 1% of the amount determined by the Tribunal to be the gross earnings of the broadcaster during the period equal to the period covered by the order that ended on the last 30 June that occurred before the period covered by the order.

(9)    PPCA, Dr Eisenach and Mr Massarsky did not propose any alternative calculation of a rate based on a percentage of revenue. Instead, PPCA’s position was that any attempt to determine a reasonable percentage of revenue rate is doomed because there is no proper evidentiary foundation to determine such a rate.

30    As a result of their entrenched positions about these fundamental matters the parties did not engage meaningfully with the opposing case. PPCA did not propose any variation to the CRA scheme (or schemes). CRA did not propose any variations to the PPCA scheme. The experts called by PPCA, as noted, provided no alternative approach to the calculation of an appropriate percentage of revenue rate. The experts called by CRA provided no alternative approach to the calculation of an appropriate per stream rate. Instead both parties maintained that there was no proper foundation for the calculation of any rate on the basis for which the other side contended.

31    After the first part of the hearing was completed in April 2015, we expressed our concern about the lack of engagement of each party with the other party’s case. We requested that the experts consider alternative approaches on the assumption we did not accept their principal approach. We also raised for the consideration of the parties a hybrid scheme, which incorporated options for payment, one being a percentage of revenue approach and one being a per stream approach. We sought the assistance of the parties as to how such a hybrid scheme might work. We provided the parties with a list of questions for the experts which included the following:

For per stream basis:

1.    What is the consequence if we do not accept that the US per stream rates are relevant/comparable for Australia?

2.    What other bases are available to determine an appropriate per stream rate if US rate is found to not be relevant/comparable (e.g. Universal Music rates for streaming or not and why)?

…

For percentage revenue basis:

1.    What is the consequence if 0.4% rate in PPCA-CRA broadcast agreement is found to be a non-market rate due to the 1% cap?

2.    What is the consequence if PPCA-CRA broadcast agreement is found to be non-comparable due to protected/unprotected difference?

…

4.    If the PPCA-CRA broadcast agreement is found to be not comparable, what other bases are available to determine a percentage gross revenue which reflects the value of the simulcast right?

5.    What are the current percentage gross revenue bases under the APRA-CRA agreement? Do you know how they were derived? Are there any other comparable agreements in this regard?

Other:

…

2.    Is a “mixed” scheme possible – e.g. (a) the greater or lesser of a per stream/percentage revenue approach; or (b) a percentage revenue approach unless a station elects to adopt a per stream approach; or (c) a per stream approach unless a station elects to adopt a percentage revenue approach?

32    Despite these requests, the entrenched position of each party remained. Neither party made any attempt to identify how a scheme embodying alternative options for payment might operate. Nor did they propose alternative rates which they considered might be reasonable if their primary approach was rejected by the Tribunal. Mr Samuel alone provided a further alternative approach but his further alternative did not engage with the possibility that a per stream payment might be appropriate. Instead, Mr Samuel proposed an additional method by which a percentage of revenue approach could be adopted, albeit not using the 0.4% rate fixed in the PPCA-CRA Broadcast Agreement as his starting point.

33    What the parties refused to recognise is that their positions are self-defeating. If CRA is right, then the Tribunal cannot properly determine any per stream rate. If PPCA is right, then the Tribunal cannot properly determine any percentage of revenue rate. Accordingly, if both are right, then the Tribunal is precluded from doing anything other than rejecting all of the schemes as unreasonable. The parties would be left with no resolution, the present matter having been an exercise in futility involving enormous wasted time, effort and costs.

34    The Tribunal has been left in an unsatisfactory position. In particular, it does not have the benefit of PPCA’s position if a percentage of revenue rate is considered appropriate. It also does not have the benefit of CRA’s position if a per stream rate is considered appropriate. It has been confronted with two separate cases and, despite its requests, has been left with the arguments for one side only for each case. Given that proceedings before the Tribunal are meant to be administrative rather than adversarial in nature (see s 164 of the Copyright Act), the intransigence of the parties is difficult to understand. It suggests the development of a forensic culture unsuited to the Tribunal’s function which should be required to change.

35    The extent of the Tribunal’s power to vary a scheme was raised during the hearing. It will be recalled that s 154(4) refers to the Tribunal “confirming or varying the scheme or substituting for the scheme another scheme proposed by one of the parties”. The Tribunal is not given by this section the power to determine a scheme it considers would be reasonable in the circumstances having regard to the schemes referred to it by the parties, which is regrettable. Provided that the requirements of procedural fairness are satisfied, we can see no reason why the Tribunal should be confined to confirming or varying a scheme referred to it by the parties, as opposed to a power to determine a scheme which it considers reasonable in the circumstances. Such an amendment would also better reflect the administrative nature of proceedings before the Tribunal. We, however, are bound by the existing legislation.

36    In the Nightclubs case the Full Tribunal said:

Section 154(4) permits the making of variations, in the sense of amendments and alterations to a proposed scheme, but does not empower the Tribunal to substitute a scheme of an entirely different kind (Australasian Performing Rights Association Ltd; Re, ABC (1985) 5 IPR 449 at 458–9).

37    In Reference by Phonographic Performance Company of Australia Ltd under section 154(1) of the Copyright Act 1968 [2010] ACopyT 1; (2010) 87 IPR 148 (the Gyms case) the Tribunal said:

[285]…The Tribunal has considered its jurisdiction under s 154(4) of the Copyright Act on several occasions, including in the ABC case (Copyright Agency Ltd v Dept of Education (NSW) (1985) 5 IPR 449). In that case, the Tribunal considered whether s 154(4) enabled it to substitute a new scheme for one proposed that it considered to be totally unreasonable. The Tribunal referred to the contrast between the language used in ss 148–151 on the one hand, where the Tribunal’s task is to determine “equitable remuneration” and the reference to variation of the scheme in s 154 on the other hand. The Tribunal concluded that the word “variation” was not to be construed so widely that it would empower the Tribunal to substitute for the scheme, which is referred under the section, a scheme of an entirely different kind in cases where the Tribunal concluded that the referred scheme was wholly unreasonable. The Tribunal considered that in such a case the only course was to make no order on the application. The Tribunal held, however, that it could vary the percentage of expenditure upon which the proposed formula was based or vary the base upon which the calculation was to be made.

[286] The 2006 amendments addressed the problem identified in the ABC case. The Tribunal does not consider that it was Parliament’s intention to confine the Tribunal’s powers to adopting unchanged such an alternative scheme. It would be an odd result if the Tribunal could vary a scheme proposed by PPCA in its reference but could not vary a scheme proposed by Fitness Australia as an alternative. At the very least, the Tribunal has the power to fix the appropriate rate for the use of copyright material and is not bound to accept a rate proposed by one or other of the parties in a proposed licence scheme.

38    PPCA initially submitted that a hybrid scheme would be beyond the scope of the Tribunal’s power of variation in s 154(4). PPCA, however, withdrew this submission and relied instead on a submission that there was no proper foundation in the material to determine a reasonable percentage of revenue rate. CRA did not submit that the implementation of a hybrid scheme, incorporating both a percentage of revenue rate as CRA had proposed and a per stream rate as PPCA had proposed, would be beyond power, but it maintained that there was no proper foundation in the material to determine a reasonable per stream rate. As we have noted, if both are right in this regard, then the inevitable result is that all the schemes referred to the Tribunal must be rejected as unreasonable.

39    Certain factors suggest that a hybrid scheme would be within the Tribunal’s power. Where used in statute the singular includes the plural and the plural includes the singular (s 23 of the Acts Interpretation Act 1901 (Cth)). Accordingly, the Tribunal may have referred to it in the one reference under s 154(4) any number of schemes and is bound to consider each such scheme. The Tribunal can confirm or vary “the scheme” which must mean any one or more of the schemes which has been referred to it. The Tribunal can thus confirm or vary multiple schemes but, no doubt, in so doing must clarify how they are to work together. If we are wrong about this then the spectre of rejection of all of the schemes arises. Standing alone, we have concluded that no scheme referred to us is reasonable in the circumstances. Because, as will become apparent, there are other issues which arise about the Tribunal’s power to vary a scheme or schemes in this matter, as well as issues about procedural fairness, we propose to invite the parties to consider their positions once they have had an opportunity to review these reasons.

1.3    Consequences of the approach of the parties

40    The unhelpful approach of the parties has had a number of practical consequences.

41    The first consequence is that, as might be expected in a case involving such deep divisions between the parties, we found the case of each party unpersuasive in significant respects, the cases reflecting the extreme position of each propounding party, resolute in an apparent effort not to contemplate the possibility of a different perspective.

42    The second consequence is that because of the lack of any proposed alternatives by both parties should their primary cases be found not to be reasonable, judicial estimation has a greater role than might otherwise be the case.

43    Third, there are consequences for the notional bargain that the Tribunal is capable of conceiving could be struck between these parties.

44    This third point, about the notional bargain, requires further explanation. It will be recalled that the only assumption that is made for this purpose is that the parties are willing to negotiate and conclude a bargain. Otherwise, the parties are taken as they are, with whatever strengths and weaknesses in their bargaining power the available material discloses they possess.

45    One piece of evidence, insignificant on its own but important when considering the notional bargaining construct, was that of Anthony Clark, the former director of Phonographic Performance Limited (PPL) in the United Kingdom (the UK). PPL is the equivalent to PPCA. The CRCA, referred to in Mr Clark’s evidence, is the UK equivalent to CRA. He gave evidence in the following exchange:

So far as you’re aware, is it the case in the United Kingdom that each time PPL has sought to negotiate or have reviewed the radio station’s use of the broadcast right, that review or negotiation has always been conducted on the basis that the percentage of revenue is the appropriate structure for the use of the sound recordings?---No. That is not the case. Certainly, the documentation I reviewed relating to the negotiations between PPL and the CRCA as it then was were looking to apply a different formula. In particular, a consumption-based formula, generally referred to as a rate-protracted stream, and they were unsuccessful in that. And I myself, when I – subsequent to my arrival at PPL, and when I also undertook negotiations with the commercial radio industry in respect of the online rights, we were – did seek to apply a consumption-based formula, but the radio industry was unwilling to contemplate that formula, so we were unsuccessful in those negotiations as appears from my look at the records of – that my predecessors at PPL were similarly unsuccessful.

46    Mr Clark expanded on what had happened in these statements:

•    And it was apparent, from my looking at those, that PPL sought to licence internet simulcast on a different basis but all its attempts to do so basically – a consumption basis or a rate per track – per stream formula. But all those attempts were resisted by the radio industry and, as at that time, it’s apparent that the extent of simulcasting – its significance was relatively minimal – a decision was taken by PPL and/or its stakeholders not to – to basically concede the position taken by the radio broadcasters on this issue.

•    As I say, it was before I got to PPL but I have looked at the papers before giving my affidavits and it was apparent that it was not the beginning of negotiations, which I think go back to 1999 when they first started talking about the basis on which internet simulcast could be licensed. And the original position adopted by PPL in these discussions was not a share of revenue. But they – as happens in negotiations, they lost that point.

•    … they were concerned that failure on the part of PPL and the radio industry to reach an agreement on this point would result in a – in the issue, along with a range of other issues going back to the Copyright Tribunal for determination, which was not something that PPL or its stakeholders found appealing at that point. So there was a whole range of issues which led to them conceding the position of the radio industry, which the – which was that the share of revenue formula which had emerged from the 1993 tribunal decision should be extended to apply to internet simulcasts.

•    Those negotiations were conducted by myself on behalf of PPL, so I am familiar with that process. It’s correct to say that from the very outset commercial radio were unwilling to enter – even enter any kind of dialogue with PPL which involved a change in the share of revenue formula. So at a relatively early stage in that dialogue, in the interests of making some progress, and at least reaching agreement on some other aspects of licensing, which I took the view along with my colleagues and my stakeholders at PPL, that if we were to make any progress with at least modernising and bringing up to date some aspects of the licence, reopening the issue of the underlying payment formula – the share of revenue which were those rates going back to 1993 – we were going to get nowhere, and the only place to go if we wanted to take that issue up was back to the tribunal. There was no appetite for that on behalf of the stakeholders. There were probably more pressing and more significant issues to address at that time, and therefore that series of negotiations which led to the 2006 negotiations did not reopen or change the share of revenue formula, because the other side were unwilling to contemplate that.

•    There were – it was – it was one of a number of issues, but it was very apparent that we were going to – if we wanted to change the revenue formula or the revenue structure, or the rates, the only place we could do that would not be through negotiation but through a tribunal. That was made very clear on behalf of the radio industry, and I understood that. PPLs decision was there were areas which we wanted to progress, we wanted to try and resolve, and we progressed those, and accepted that we were not going to – unless we were prepared to go to a tribunal, we were not going to reopen the revenue share formula at that stage.

47    In other words, on more than one occasion, PPL had sought to negotiate payments based on a per stream approach but the UK equivalent to CRA refused to conduct negotiations on that basis.

48    In the present case the material made available indicates that in any actual negotiation between the parties (which is what the notional bargain is attempting to replicate) each side would consider certain matters to be fundamental. In summary, we consider it apparent from a broad overview of all of the material that:

(1)    CRA would not negotiate an agreement with PPCA for stations to be licensed to simulcast by which the stations would be required to pay on a per stream rate. CRA considers that a requirement to pay on a per stream basis would be catastrophic for its members irrespective of the amount of the rate actually imposed.

Although we are satisfied that CRA’s belief that a per stream approach would be a catastrophe is irrational, in any negotiation CRA would have available to it other persuasive and rational arguments in favour of a percentage of revenue rate. In particular, history and circumstance weigh in favour of a percentage of revenue rate. CRA’s members have not to date been charged on any basis by PPCA (or any other collecting society including APRA) other than a percentage of revenue. CRA’s members have in place systems and procedures which enable them to pay licensing fees for the use of music on a percentage of revenue basis. A per stream scheme for the simulcast right would result in CRA’s members having to implement two systems relating to the transmission of the one radio program, the applicable payment system being dependent on the method of transmission (broadcast over the air or simulcast over the internet). Whatever the future potential for monetisation of simulcasting, Australian radio stations currently make the vast majority of their revenue from advertising which, presently at least, is based on total audience numbers irrespective of the mechanism by which those listeners are reached. Further, none of the problems with assessing revenue and gaming the system identified by PPCA have been thought sufficient to undermine the percentage of revenue approach to date.

All of these matters would place CRA in a relatively strong bargaining position with PPCA about the structure of the rate.

(2)    PPCA would have available its own cogent and persuasive arguments that a per stream rate is an appropriate method of payment for the activity of simulcasting, providing significant advantages when compared to a percentage of revenue approach. Simulcasting, unlike broadcasting, enables a per stream approach to be adopted. A per stream approach means that stations pay for their actual music use. Stations can thus control the costs of this input by adjusting their music use. A per stream approach rightly recognises music as an input cost to the product of the radio stations. A per stream approach avoids difficulties of assessing revenue and gaming of the payment system. The spectre of technical impossibilities or difficulties and prohibitively expensive compliance costs which appear to inform CRA’s belief that the imposition of a per stream rate would be a catastrophe are not supported by any rational approach to the available material. That this is so is demonstrated by the fact that in the US the body equivalent to CRA is not seeking to abandon the per stream approach in the latest round of hearings. It is seeking only a substantial reduction of the rate per stream. If a per stream approach is truly impractical and unworkable, or extraordinarily onerous from a technical or compliance respective (as CRA would have it), then it could be expected that in the US CRA’s equivalent would not be seeking a rate reduction; it would be seeking a different payment system altogether.

All of these matters, except perhaps the gaming concern where CRA’s answer would be equally valid, would place PPCA in a relatively strong bargaining position with CRA about the structure of the rate.

(3)    PPCA would not agree to a percentage of revenue rate which was based on the 0.4% rate in the PPCA-CRA Broadcast Agreement. It would not do so because (rightly, in our view) it would judge that rate to be largely the result of the operation of s 152(8) of the Copyright Act and the 1% cap it imposes, rather than a market rate. While CRA would make arguments to the contrary none of them (as we have found below) would be persuasive. PPCA would be in a strong position to maintain that the 0.4% from the PPCA-CRA Broadcast Agreement is an inappropriate starting point because it does not reflect a market rate for that right and, accordingly, cannot be a proper basis for a market rate for the simulcast right. CRA would not be in a strong position to support a percentage of revenue rate starting from the base of 0.4%. PPCA’s position that the 0.4% is not an appropriate staring point would be based on far better ground.

(4)    CRA would not agree to a per stream rate based on the rates in the NAB-SX agreement or webcasting agreements (that is, agreements between record companies and music streaming services). CRA would be in a better position than PPCA on this issue because of certain incontrovertible facts. First, the US commercial radio market is far larger than that in Australia. Second, it is likely that cost structures in the US enjoy benefits of economies of scale that are not present in the Australian market to the same extent. As such, costs of implementation and ongoing compliance of a per stream scheme in Australia are likely to be greater than in the US. Third, there is no protection for use of sound recordings in radio broadcasts in the US so that the radio stations are paying only for simulcasting and not for broadcasting. The effects on the US market if they also had to pay for broadcasting rights are unknown. Fourth, the US is further advanced than Australia in terms of the monetisation of simulcasting. Fifth, the rate reduction being sought in the US to the rates set by the NAB-SX agreement is substantial, a reduction from $0.0025 per stream (or 0.25 cents) to $0.0005 per stream (or 0.05 cents).

Insofar as webcasting agreements to enable music streaming services are concerned, all provide a greater interactivity than the simulcasting of a radio program (even allowing for substitution of advertisements which PPCA will accept) and all provide a different experience from the radio stations which CRA represents. While there are some very high music use stations, other stations involve low music use (sports and talk back stations), and the majority provide music to listeners as part of an overall package of entertainment of which music forms one part (albeit, in the case of some stations, an important part). The same cannot be said of webcasters which simply stream music. The presence of advertisements in some webcasts does not make those services particularly like radio given the package entertainment nature of radio. Moreover, commercial radio in Australia has a long history, unlike webcasting, and throughout its history it has never been subscription based. It is funded by advertising revenue. The business models of webcasters are different, their bargaining position as against record companies is different as their business is wholly dependent on music, and they could be expected to pay more for that product than radio stations.

(5)    Both parties would be aware of the fact that many contemporary agreements between record companies and webcasters provided for payment on a percentage of revenue or per stream basis whichever is the greater. Whether one or other calculation would never be used in practice because of the nature of the webcasting business as PPCA submitted (that is, that payments would always be made on the per stream basis because this will always be greater than the revenue earned given the nature of the webcasting business) is not to the point. The point is that it is commonplace for copyright owners and those who wish to use music to agree that payments should be made on both percentage of revenue and per stream bases. In this context, it is unlikely that in any notional bargain between PPCA and CRA consideration would not be given to a licence which provided alternative bases for payment.

(6)    PPCA would not agree to a scheme which gave to CRA any greater role than practically necessary or, preferably, gave to CRA any role at all. On PPCA’s side in this regard would be the fact that membership of CRA is not compulsory. Further, PPCA is not granting a licence to CRA but to the radio stations which wish to simulcast. Rightly or wrongly, PPCA also perceives CRA to have been difficult to deal with in respect of the PPCA-CRA Broadcast Agreement. On CRA’s side would be the fact that it currently represents a large number of radio stations. It has a current role as the contact point with PPCA under the PPCA-CRA Broadcast Agreement. It is also likely that even if the scheme minimises CRA’s formal role, radio stations are likely to seek assistance from CRA in complying with their obligations in any event.

It seems likely that in any notional bargain a practical solution to this impasse would be able to be reached whereby CRA’s role was only so great as any particular radio station might allow provided that the obligations remained those of the radio station directly to PPCA, rather than the radio stations to CRA and CRA to PPCA.

49    We do not suggest that our conclusions about the likely actual bargaining position of the parties are conclusive of the Tribunal’s task. If that were so, then inferred bargaining positions based on unsubstantiated concerns, such as those of CRA about the practical difficulties of implementing a per stream approach or those of PPCA about the likelihood of gaming of a percentage of revenue approach being an insuperable problem, might carry undue weight. One of the advantages of the notional bargaining concept is that it can be assumed that, ultimately, arm’s length commercial negotiations are likely to yield an agreement which is based on genuine, rational, and well-founded positions. But in the context of the Tribunal’s task of confirming or varying a scheme or schemes as it considers reasonable in the circumstances, the position likely to be taken by the parties during the course of any actual negotiation is not irrelevant. It does not determine conclusions about the bargain that ultimately would have been reached, but it does inform the Tribunal’s assessment of what is reasonable in the circumstances.

50    These considerations indicate that confirming any one scheme would not be reasonable. Rather, as we suggested to the parties after the first part of the hearing was completed, they indicate that a hybrid scheme adopting varied per stream rates and varied percentage of revenue rates is likely to be reasonable in the circumstances.

51    We explain our reasons for these conclusions and provide greater detail of our conclusions thus far in the following sections.

2.    OVERVIEW OF THE EVIDENCE

52    We provide a brief overview only of the most important evidence at this stage because, as we have said, much was peripheral or dealt with imponderable matters which, if necessary to resolve at all, are ultimately to be resolved at a level of impression or estimation. This overview is based on summaries provided by the parties.

53    Joanna Dick is the head of Off-Screen Media Direction Pty Ltd (OMD) and in that role is responsible for booking all outdoor, radio and print advertising for OMD Australia’s clients. CRA called Ms Dick to address the buying and selling of advertising on radio from a media buyer’s perspective.

54    Deborah Hishon is the head of Client Service, Radio at GfK Retail and Technology Australia Pty Ltd, the official provider of radio ratings in Australia. CRA called Ms Hishon to address the methodology and process of radio ratings audience measurements and figures in Australia for the commercial radio industry.

55    Eve McGregor is the head of Legal and Regulatory Affairs for CRA. CRA called Ms McGregor to provide an overview of CRA, its regulatory structures, the licence fee agreements and reporting obligations.

56    Paul Bowd is responsible for agency sales for Macquarie Radio Network’s two metropolitan stations, 2GB and 2CH. CRA called Mr Bowd to provide an overview of the metropolitan business of Macquarie Radio Network.

57    Ronald Camplin is the executive chairman of Bathurst Broadcasters which owns the commercial radio stations 2BS AM and B-Rock FM. CRA called Mr Camplin to provide an overview of the business of Bathurst Broadcasters, being the operation of small regional stations.

58    Geraint Davies was the chief operating officer of Australian Radio Network (ARN) and is now the head of iHeartRadio. CRA called Mr Davies to provide an overview of the business of ARN as a large metropolitan radio network, including the introduction of iHeart Radio.

59    Catherine O’Connor is the CEO of Nova Entertainment Pty Ltd. CRA called Ms O’Connor to provide an overview of the business of Nova, a large metropolitan radio network.

60    Stephen Everett is the Managing Director of ACE Radio. CRA called Mr Everett to provide an overview of the business of ACE Radio, a regional network.

61    Kevin Blyton is the managing director and joint owner of Capital Radio Network, which operates a number of radio stations around the ACT and regional southern New South Wales, as well as Perth. CRA called Mr Blyton to provide an overview of this regional network.

62    Nick Morgan is the digital manager of West Coast Radio. CRA called Mr Morgan to discuss the reporting proposed by PPCA in the context of the business of this Western Australian regional network

63    Antony Samuel is a forensic accountant and valuation expert with Sapere Research Group. CRA called Mr Samuel to provide expert evidence as to an appropriate structure and rate for the simulcast scheme.

64    Roy Epstein is a consulting economist and Adjunct Professor of Finance at the Carroll Scheme of Management, Boston College. CRA called Dr Epstein to provide expert evidence as to an appropriate structure and rate for the simulcast scheme.

65    Matthew Payton is the director of external affairs at RadioCentre Ltd, UK. Mr Payton’s evidence, adduced by CRA, provided an overview of the commercial radio market, simulcasting and copyright fees paid in the UK.

66    Gabriel Van Loon is a partner in the law firm Van Loon Simmons Professional Corporation in Canada. Mr Van Loon’s evidence, adduced by CRA, provides an overview of Canadian commercial radio and webcasting markets, copyright and broadcasting law, the nature of simulcasting and collective administration.

67    David Innes is a broadcasting and marketing communications consultant in New Zealand. Mr Innes’ evidence, adduced by CRA, provides details of the Tribunal proceedings in New Zealand which gave rise to the current scheme in that country.

68    Mark Garnett is a forensic technology expert. Mr Garnett’s evidence, adduced by CRA, provides details of the IT requirements necessary to comply with PPCA’s proposed final scheme.

69    Lynne Small is the General Manager of PPCA. PPCA called Ms Small to provide evidence of PPCA’s membership, repertoire and distributions, the history of the litigation regarding the simulcast right and the 1% cap, and PPCA’s reasons for proposing the PPCA Scheme.

70    Karen Don is the General Manager, Legal and Business Affairs of Universal Music Australia (Universal). PPCA called Ms Don to explain the increasing significance of revenue generated through the online use of sound recording and how the internet has changed record companies’ business model and the importance of licensing revenue. Ms Don also gave evidence about and arranged for Universal to provide confidential information about its recent agreements with webcasters.

71    Michael Taylor is the Managing Director of Universal Music Australia. PPCA called Mr Taylor to give evidence of the record companies’ significant investment in the creation of sound recordings, via the “Artist and Repertoire” process, including scouting, developing, managing, supporting and promoting recording artists, and the high risks associated with that process.

72    Brent Coker is a Lecturer of Marketing at the University of Melbourne and the managing director of Deloosh, a technology company that specialises in the development of web applications (or apps). PPCA called Mr Coker to explain the opportunities for radio stations by simulcasting and monetisation strategies for online businesses.

73    Harry Emerson III is a co-founder, partner and senior executive of SurferNETWORK, an internet media streaming business in the United States. Mr Emerson’s evidence, adduced by PPCA, explained the royalty reporting services that SurferNETWORK offers to radio stations and the feasibility of royalty reporting for simulcasts in Australia.

74    James Kott is senior vice president of products and marketing at Abacast, a provider of streaming, live and on demand ad insertion and monetisation solutions for online radio in the United States. Mr Kott’s evidence, adduced by PPCA, explained the advantages of internet radio and strategies for effective monetisation of simulcasts.

75    Jeffrey Eisenach is senior vice president and co-chair of the communications, media and internet practice at NERA Economic Consulting. PPCA called Dr Eisenach to provide expert opinion evidence in support of PPCA’s scheme, including as to the structure and amount of the rate.

76    Barry Massarsky is an economist specialising in copyright issues in the music industry. PPCA called Mr Massarsky to explain the economic underpinnings of the per stream rate in the US.

77    Rodney McKemmish is a partner and the head of forensic technology at PPB Advisory, a firm offering accounting, forensic and reconstruction services. Mr McKemmish’s evidence, adduced by PPCA, describes options available to radio stations to provide per stream reporting information from a technical standpoint.

78    Anthony Clark was the director of licensing at PPL in the UK. PPCA called Mr Clark to provide evidence regarding how radio broadcasting and simulcasting is licensed in the UK, and the difficulties experienced with the implementation of a percentage revenue scheme in that jurisdiction.

79    Damien Vaughan is the CEO of Recorded Music NZ. Mr Vaughan’s evidence, adduced by PPCA, explains how radio broadcasting and simulcasting is licensed in New Zealand and describes the developments that have occurred since the most recent decision of the NZ Copyright Tribunal in relation to the licensing of radio simulcasts.

80    It is appropriate to say something here about the four main expert witnesses, Dr Eisenach, Mr Massarsky, Dr Epstein and Mr Samuel. Each party criticised the experts who expressed opinions contrary to the case the party wished to advance on the basis not only of the cogency of their opinions, but matters relevant to overall credibility. We did not find the criticisms relating to the overall credibility of these witnesses helpful. Each witness disclosed their experience and background and what they had done. We consider each was doing his best to ensure we understood the bases for the conclusions reached.

81    In addition, numerous volumes of documents and exhibits were tendered to support the competing propositions of the parties, in all, over 13,000 pages of material, mostly annexures and exhibits to affidavits, much of which was not ultimately referred to by the parties and added nothing to the weight of their respective cases.

3.    THE STRUCTURE OF THE RATE

3.1    Overview

82    In its opening submissions PPCA made the following points in support of a per stream approach:

(a)    The activity concerned is being conducted entirely over the internet.

(b)    The internet is – uniquely, considered historically – adapted to yield, at no or very low cost, near-perfect census information about usage.

(c)    Census-based usage – i.e., per stream, per listener – is inherently fair both to the copyright owner (represented here by PPCA) and to the copyright user (the radio station) in that, once a price has been fixed for each stream, the amount payable rises or falls with the actual usage of the copyright subject-matter concerned, here, sound recordings.

(d)    When compared to a percentage of revenue rate (as contended for by CRA), a census-based rate also ensures that royalties are apportioned fairly between users, as each station pays for its actual usage, rather than an arbitrary percentage of revenue that does not vary between stations that have a significant portion of their audience streaming and significant online revenue, and those that make very limited use of the simulcast right.

83    In its closing submissions PPCA said:

it cannot sensibly be disputed that – all other factors being equal – a rate that is closely tied to the actual extent of use of the copyright subject-matter in the activity in question will be more reasonable and equitable than one that is not.

84    We accept this submission for numerous reasons as set out below. What we do not accept, however, is that all other factors are equal. As discussed above, history and circumstance, which are important factors, weigh against the imposition on radio stations of a requirement that they be forced to implement a new system for payment of fees for use of sound recordings in simulcasts of their radio broadcasts given that they are already paying for the use of the sound recordings in their broadcasts on a percentage of revenue basis. We do not say this because of any concern about double counting. If there is double counting (discussed below), because radio stations pay on total gross revenue (which includes revenue derived from simulcasting), an adjustment could be made to address this issue. We say this because, unlike the position in the US, there is a long history in Australia of radio stations paying for use of protected sound recordings in radio broadcasts on a percentage of revenue basis and, irrespective of this matter, they will continue to do so for the foreseeable future.

85    CRA rightly made these points in its submissions. It referred to the Tribunal having adopted a percentage of revenue approach in the 2MMM case. It referred also to other contexts in which a percentage of revenue approach has been adopted including what the Tribunal said in Reference by Australasian Performing Right Association Ltd; Re Australian Broadcasting Corporation (1985) 5 IPR 449 at 479 as follows:

A percentage of revenue has a long history of acceptance as a measure of the worth of copyright. It is accepted in the area of literary works where the owners of copyright in books receive a percentage of retail sales. It also applies in the case of sales of sheet music where again percentages of sales are taken as a guide to what the copyright owner should receive. And it is accepted in the theatre and concert areas where copyright owners receive a percentage of the box office receipts. The philosophy underlying this approach… [is] that over the years the copyright owner has been perceived to have an interest in the success or otherwise of his work. If it is highly successful and substantial returns are yielded, he should receive more. If his work is a failure, he will receive little or nothing. This approach has particular application in commercial cases, that is where the licensee, in this area of copyright use, the broadcaster, is a commercial entity.

86    In Reference by Australasian Performing Right Association Ltd under s 154 of the Copyright Act 1968 (1992) 25 IPR 257 a similar approach was adopted, the Tribunal saying at 266 that:

The box office or gross revenue approach has therefore the solid basis of past experience in a variety of areas to commend it.

87    At 267 the Tribunal said:

In the end the ultimate matter the Tribunal has to consider is the reasonableness or otherwise of the amount. The method or base which is used to calculate it is important because the selection of a proper method is likely to yield the fairest result. But the result which is thus yielded must be considered against the background of the past, in particular the level of fees payable down to the present time. If the adoption of the amount achieved would suggest an extravagant outcome, the matter needs to be reconsidered.

88    We accept that, unlike the position in the US when per stream rates were adopted for the simulcasting of radio broadcasts, we do not have the freedom of a blank slate. In circumstances where radio has paid for the use of sound recordings in broadcasts on a percentage of revenue basis and will continue to do so for the foreseeable future, to impose a scheme which requires all radio stations to pay for the use of sound recordings on a per stream basis, in our view, would not be reasonable in the circumstances.

89    That said, if, as in the US, we were confronted with the need to determine a reasonable scheme for payment for the use of sound recordings in simulcasts where there was no arrangement already in place for payment for use of sound recordings in broadcasts, then the advantages of a per stream approach would be overwhelming and no consideration would need to be given to any role for a percentage of revenue approach.

90    We say this for the following reasons.

3.2    The technical and practical issues

91    We found the evidence adduced by the radio stations about the technical and practical difficulties with implementing a per stream approach, and the impacts it would have on their freedom of decision in respect of choice of play-out systems, technology upgrades and use of third parties for various parts of the process, unpersuasive. This evidence failed to recognise a few basic facts.

92    First, not only have radio stations managed to operate under such a system in the US for years, but also having been given the opportunity to present a case about the system in the US (as is the case in the present Webcasting IV proceedings in the US concerning the rates and terms that should be imposed for the period 2016 to 2020) the radio stations are not suggesting that a different approach should be implemented. Their case is that the current rates they are required to pay are too high and the reporting requirements should be amended. If any of the technical or practical difficulties presented in the large amount of evidence adduced by the radio stations in this matter were incapable of resolution at a reasonable cost and in a reasonable manner, then it is difficult to understand why US radio stations, which have actual experience of the system, would not be advocating for the per stream approach to be abandoned. The fact that they are not doing so effectively answers all of the evidence which CRA adduced about these issues.

93    Second, the technical and practical feasibility of a per stream approach is also supported by the fact that it is the dominant method for payment in respect of all webcasting agreements which found their way into evidence. We accept that the webcasting business is not the same as the commercial radio business. But that is not the point. The point is that none of the webcasters around the world appear to have had any difficulty in negotiating an agreement under which they pay for their use of sound recordings on a per stream basis. The obvious inference from this fact is that such a payment scheme is technically feasible and practically able to be implemented. If webcasters believed that such a payment system would involve intolerable burdens of cost and time, or an unreasonable restraint on their commercial decision-making, then it could be expected that they would not have agreed to make payments on this basis. Yet such agreements appear to be common-place.

94    Third, it is apparent that when a new need arises the market responds to fulfil that need. As PPCA noted, there was no history of per stream reporting by the US radio industry when the US Copyright Royalty Board ordered it in the Webcasting I decision in 2002. Yet when payments were required to be made on a per stream basis the companies which provided streaming services to radio stations expanded their business to enable compliance with the requirements for payments on a per stream basis. Whatever the complaints which might have been made about that system and the burdens it imposes the fact remains that now, more than a decade later, the radio stations are not suggesting in the Webcasting IV proceedings that the system is unworkable or should be changed to one providing for payment on some other basis. If the US experience had been the calamity that CRA would have it then this fact is inexplicable.

95    Further, while it might be expected that Mr Emerson, a founding partner of the streaming provider SurferNETWORK, would present his business in the best possible light, the other inescapable fact is that SurferNETWORK is a functioning business which provides the necessary services in the US to enable radio stations to comply with their obligations to make payment on a per stream basis. We have no doubt that such services would become available in Australia at a cost which might be greater than that in the US due to the different size of the industry but would nevertheless be commercially acceptable to any radio station which is commercially viable.

96    Equally, we have no doubt that the kinds of problems CRA and the witnesses it called emphasised have all been experienced and satisfactorily resolved in the US. By satisfactorily resolved we do not mean that the system is perfect or beyond complaint. This was another fallacy in CRA’s approach to the issue of a scheme on a per stream basis. While inexactitudes and inaccuracies inherent in payments on a percentage of revenue basis were seen as acceptable by CRA, it focused what can only be described as extraordinary energy on identifying and dissecting every possible inexactitude and inaccuracy inherent in a per stream basis payment system. The reality is neither system will be perfect. If, however, ensuring that an individual station pays what it should and not more or less is a valid objective (and we think it is), then a per stream payment system is better adapted than a percentage of revenue payment system to achieve that end. None of the potential problems which were identified are insoluble (as the arrangements in the US disclose) and none were shown to have any potential impact on the accuracy of the system which would be greater than the uncertainties and estimations necessarily involved in a percentage of revenue payment system.

97    Fourth, we found the evidence of Mr Garnett, a computer expert called by CRA, unpersuasive. Mr Garnett appeared to place no weight on the fact that every issue he raised must have been resolved to a satisfactory extent for the US system to operate and for the US radio stations not to be seeking a radical overhaul of that system. His opinions were thus based on speculation about what might happen rather than the reality of the US experience (and the experience of webcasters). Mr McKemmish’s evidence, to the contrary, was inherently credible and persuasive. As PPCA put it in respect of the primary concerns which Mr Garnett raised:

(a)    Robotic connections/non-genuine connections – Mr Garnett raised this as a “technical issue”, although it is in truth a reporting or classification issue. Mr McKemmish identified straightforward solutions to this issue. Mr McKemmish proposed simple solutions to these matters. Geo-blocking was one technique discussed. Notoriously, robotic services and the like are not based in Australia: indeed, Mr Morgan’s evidence was all of overseas monitoring services. Many of Mr Garnett’s concerns disappear. Moreover, this is an issue squarely addressed by Mr Emerson: in the US, asked and answered. But notably these “non genuine” users are not of concern in practice in the US: of SurferNETWORK’s 400 radio station customers, to whom the pop-up reminder button is made available, only about 6 use the feature in practice. Mr Emerson’s radio station customers are not concerned about this issue.

(b)    Syndicated shows – Mr Garnett saw this as a “difficult question”. He said:

… it begs the question, how does the radio station know what music is being played during the course of that show.

He described it as “another degree of complexity”. However, this issue has been faced and dealt with in the US.

(c)    Minimum connection times – the experts debated this; again it came down to an issue of tweaking or filtering the systems. But again, this question has been confronted in the US and dealt with. And this is par excellence something that is likely, in short order, to balance out over- and under-reporting. Moreover, any filter could be set to filter out not only connections to thirty seconds, but also song-plays less than, e.g., fifteen seconds.

98    In any event, as we have also said, the notion that any system will enable complete accuracy is unrealistic. Every payment system will involve elements of estimation and imprecision. The idea that a payment system based on a percentage of revenue is not itself subject to these kind of imperfections, which seemed to be CRA’s position, is also unrealistic. Indeed, by its very nature a percentage of revenue payment system necessarily involves some radio stations paying too much and some paying too little for their music use because the key driver of the calculation is not music use at all. Insofar as music use is factored in, a broad brush approach is taken to categories of music use percentages. The kinds of inaccuracies which are inherent in the per stream payment system are of a lesser order of magnitude than those inherent in a percentage of revenue payment system.

99    By way of example, CRA submitted that because the PPCA scheme involves a minimum connection time of 30 seconds:

a user who is listening online and flicks between a number of songs on a number of stations may generate a royalty fee payable to PPCA for each song he or she listens to, even where that listening may last no more than a second or two provided that the connection reaches 30 seconds.

100    In contrast, listening for less than eight minutes in a 15 minute period is not counted for ratings purposes. CRA said:

A radio station should not have to pay for that which has no value to it. And PPCA should not receive a royalty of more than it would be entitled for damages for copyright infringement (and it would be entitled to none on a stream of a track of a few seconds or less in duration because there would be no infringement by the communication of a substantial part of the recording).

101    We do not consider this concern to be justified. As we have said, there will be imperfections in every system. CRA overlooks the fact that many connections might be 29 seconds and that these connections will not be counted as a stream at all. In this sense, as with any payment system, there are “swings and roundabouts”.

102    CRA’s submissions about these technical and practical issues, overall, were based on the same speculative and unsubstantiated fears as its evidence. The effort to implement a per stream payment system would not be “Herculean”. None of the evidence supports that characterisation. The lack of existing in-house expertise outside a few major stations is to be expected. The position must have been the same at the outset in the US. The fact that the task involved may take some time and money and is ongoing is immaterial given the evidence that the cost and time will not be unduly burdensome. The fact that many stations ceased to simulcast due to the interim scheme is also immaterial. In our view, they appear to have done so based on fears of the unknown and to bring political pressure to bear in favour of the radio industry view that simulcasting should be covered by the PPCA-CRA Broadcast Agreement. Fears of the unknown are understandable, particularly if those fears can be exploited in a policy debate. But what the evidence in this matter has shown is that, provided the rate is not excessive, the fears of a per stream payment system are unfounded. The notion that the system would mean that stations would be paying for “large numbers of hours of internet listening that did not represent genuine listeners” represents one of those unfounded fears, belied by the reality of the position in the US.

103    CRA said that the:

only evidence before the Tribunal as to the actual time and costs involved in implementing a customised per stream reporting system in a radio station environment:

(a)    ARN's reporting system involved 2-3 months of software development (equivalent to approximately $48,000-72,000) for initial implementation, 6 months of intensive validation work, primarily to address inaccuracies in interpreting the streaming logs (and aspects of that validation were still ongoing more than 12 months after initial implementation);

(b)    in the case of Nova, the 2-4 days of development and 1-2 weeks of testing to which McKemmish referred was in relation to development work of RCS, a third party vendor of its playout systems to address the issue of networked shows; this is merely a pre-requisite to the additional, and significant, amount of work to be undertaken by Nova in-house to adapt its "highly simplified" pre-existing reporting system for Coles Radio, the details of which are set out in the second O'Connor affidavit.

104    This represents a selective view of the evidence. It appears that ARN has undertaken its own work from scratch, presumably for its own reasons. Whatever ARN and Nova have done or experienced, the Australian radio stations will have the benefit of all that has been developed as a result of the US system which has operated for more than 10 years. Whatever difficulties US stations confronted when implementing a system which at that time was truly new, the issues are not such as to prevent the current operation of the system or to have caused the radio stations to contend in the Webcasting IV proceedings that the per stream payment system should be abandoned. To adopt a per stream payment system Australian radio stations will have the benefit of the US experience. It is unreasonable for CRA to expect that any payment system will be completely accurate, impose no burdens on the person required to pay, and never cause that person a moment’s difficulty. Yet that is the standard that CRA’s submissions sought to impose on a per stream payment system. We found the approach of CRA to this issue unrealistic and overblown. For these reasons we do not propose to address, on a point-by-point basis, CRA’s many further specific propositions in this regard.

3.3    A per stream approach – the advantages

105    Whatever the potential for debate about inaccuracies in the data, the object of a per stream payment system is that each radio station pays for its actual use of sound recordings. This is not the object of a percentage of revenue payment system. A percentage of revenue payment system involves a payment on a different basis – it gives the owner of the sound recordings a share of the money the radio station has managed to earn from its entertainment package of which music forms one part. While the formula for payment by each radio station on a percentage of revenue basis includes some gross indicators of music use (such as the music use percentage or MUP figures which underlie various agreements between radio stations and copyright owners), the relationship between actual use and payment in a percentage of revenue payment system is indirect and obscure.

106    Accordingly, as a matter of principle, a per stream payment system starts from what we consider is the correct premise – that radio stations use music as one input into their programs and that, insofar as reasonably practicable, they should pay for that music based on their actual use. They should not pay based on how much revenue they manage to earn from advertisers. How much money a radio station manages to earn from advertisers has nothing to do with the owners of copyright in the sound recordings. It is a product of the business capacities of the radio station and thus reflects the competencies or otherwise of those making decisions at the radio station. A radio station which excels at maximising its revenue should not pay more for the same song than one that is not so skilled, yet that is the effect of a percentage of revenue payment system. In PPCA’s words:

A licensee’s revenue is affected by many factors (for example, marketing ability, research and development, and cost efficiencies) which are well outside the control over the copyright owner. A percentage of revenue structure therefore penalises licensors for licensees’ own poor commercial decisions. By contrast, a per stream structure ties compensation directly with the use of the copyright, and is precisely related to the intrinsic value and extent of use of the music.

107    In other words, the per stream basis reflects the value that a consumer places on the product. Any other basis represents a proxy for this value.

108    It follows that we do not accept CRA’s submission as follows:

the relevant usage should be usage by the radio station in its radio programme, not usage by the listener – it is the radio station, as the licensee, which is making use of the licensed good as an input for its product (the radio programme).

109    This submission does not confront the fact that no radio station exists merely to produce radio programs. The relevant activity is providing a radio program to which people listen. In the context of simulcasting, the provision of the radio program over the internet is the use by the radio station which directly relates to the listening by a listener. This is why a per stream rate provides a more direct and transparent basis for payment than a percentage of revenue rate.

110    A per stream payment system is also inherently fair as between radio stations because it is based on actual use whereas payment in a percentage of revenue payment system is not based on actual use. If a radio station does not play much music, it will not have many streams for which to pay. If the radio station plays a lot of music, it will pay for many streams. Importantly, however, every station will pay the same amount for every stream. By its choice of use of music, therefore, every radio station starts in the same position and can control this input cost.

111    Dr Eisenach made the following points about a per stream payment system with which we agree:

(1)    Such a system ties compensation directly to usage, thereby giving the licensee appropriate economic incentives to economise on the use of the product. By contrast, licence fees under a percentage of revenue structure depend on the licensee’s effectiveness in using the licensed recordings to generate revenues, a process over which the licensor has no control.

(2)    A per performance fee directly captures the value of the specific economic activity at issue – i.e., a simulcast – to the licensee, without requiring inferences to be made about how that activity contributes to the licensee’s revenues.

112    It follows that we disagree with the opinions expressed by Dr Epstein and Mr Samuel about a per stream payment system.

113    Contrary to Dr Epstein’s opinion, Dr Eisenach was not saying that radio stations should be discouraged from using sound recordings. He was saying that stations should pay for the actual amount they use, not some other amount estimated by reference to proxies for music use.

114    Further, Dr Epstein’s opinions appeared to be founded on the principle that as economic efficiency is achieved by prices which reflect the marginal cost of an additional unit of production, it follows that the economically efficient price for the playing of an additional song on the radio is nearly zero because the additional cost of playing that song is itself nearly zero. This analysis is not persuasive. It focuses on one side of the notional bargain only – the cost to the radio stations of simulcasting (which, according to the evidence, is low). It ignores the other side of the bargain – the record companies and artists whose production costs and risks are high. As Dr Eisenach said:

The Epstein Report is ... incorrect when it ... suggests that the efficient price for the Simulcast Rate would be zero (since the marginal cost of an additional simulcast is zero)… higher rate would result in a reduction in output or cause stations to consider modifying their programming to utilize less music-oriented programs. These arguments ignore the fact that, in all but the very short run, the availability of music depends on compensation for artists and distributors that reflects the total costs of production, fixed as well as variable.

115    Dr Epstein’s opinion that a per stream payment system would involve greater economic distortions than a percentage of revenue payment system is also difficult to understand. This opinion appeared to be based on the premise that the per stream system would limit output by discouraging radio stations from playing music to a greater extent than the percentage of revenue system. No rational basis for this opinion is apparent. The opinion seems contrary to the fact that a per stream payment basis involves a direct and transparent relationship between use and payment whereas the relationship between use and payment in a percentage of revenue payment scheme is indirect and obscure. We adopt an example provided by Dr Eisenach which discloses that a percentage of revenue payment system involves greater likelihood of inequities, and thus distortions, as between radio stations than a per stream payment system. Dr Eisenach said:

the evidence on the proportion of the listening audience relying on simulcasting also supports the proposition that small regional broadcasters like Bathurst rely less heavily on simulcasting than larger metropolitan stations. Under the CRA Scheme, however, a Bathurst station - if it chose to stream even a single simulcast - would pay the same proportion of its total revenues as a metropolitan station for which simulcasting generates substantially greater value. By contrast, under the PPCA Scheme, stations like the Bathurst stations pay royalties in direct proportion to the actual number of listeners who utilize its simulcasting service.

116    Insofar as Mr Samuel is concerned, we do not accept his opinion that a per stream payment system is inappropriate because it bears no relationship to the incremental cash flows generated by simulcasting. First, the incremental cash flows generated by simulcasting bear only upon one side of the bargain, that of the radio stations. Second, Mr Samuel’s approach to the incremental value overlooks the fact that at present on average about 10% of listeners to the radio do so via the internet simulcast. Simulcasting is thus worth, on average, 10% of radio total audience share, which is by no means insubstantial in an industry which fights hard for every listener. As Dr Eisenach said if an income-based valuation method is used, the issue is what the profits would have been assuming the activity sought to be valued could not be carried out. Third, the approach overlooks the role of simulcasting in helping radio stations to maintain their audience share. Fourth, it overlooks the potential value of simulcasting to radio stations. While we accept that this potential value cannot be accurately quantified it would be naive to assume that digital technology will not make simulcasting of radio programs of more importance in the future than at present. For example, the simulcast is not geographically restricted. A radio station can provide its listeners with their listening experience anywhere in the world via the simulcast. This is likely to be a material factor in enhancing listener loyalty. Another example is the fact that, in this matter, radio stations want the capacity to substitute advertisements in simulcasts under the licence. One reason for this is obvious. Simulcasting allows advertisements to be targeted to the listening audience. Neither of these capacities is available over the broadcasting network.

117    Mr Samuel’s opinion about the inappropriateness of a per stream payment system did not grapple with the essential, and valid, point Dr Eisenach was making that the copyright owner has no role in how the radio stations choose to make money and thus should not be required to financially participate in the consequences of the choices made by the radio station as to how they use the simulcast right. As Dr Eisenach said, a per stream payment system, unlike a percentage of revenue payment system, places “the incentives for putting the licence to its highest valued use where they properly belong, on the broadcasters”.

118    We also do not accept the concern about double counting. It was CRA’s position, based on the evidence of Dr Epstein and Mr Samuel, that a per stream payment system necessarily involves double counting. By this, what is meant is that under the PPCA-CRA Broadcast Agreement radio stations pay based on their total revenue, which includes any revenue attributable to simulcasting. If they also pay for the right to simulcast on a per stream approach, according to this argument, they will be paying twice for the same right. As CRA put it:

The Industry Broadcast Agreement requires radio stations to pay (and continue to pay) 0.4% of all gross revenue (regardless of whether it is derived from advertising in radio programmes broadcast or simulcast. If a per-stream structure is adopted for simulcasting, double counting will continue to occur.

119    The problem with this argument is that it confuses payment for the right to use sound recordings in the simulcast with the method of calculation which applies to payment for the right to use sound recordings in broadcasts required by the PPCA-CRA Broadcast Agreement. The parties, and the Tribunal, are bound by the decision of the Full Court which determined that the PPCA-CRA Broadcast Agreement does not apply to the simulcast right. On that basis, the radio stations, in accordance with the PPCA-CRA Broadcast Agreement, are paying for the broadcast right only. It happens that the PPCA-CRA Broadcast Agreement provides that payment for that right shall be on a percentage of revenue basis with the relevant revenue being total revenue. This does not mean, however, that the payment under the PPCA-CRA Broadcast Agreement is also a payment for the simulcast right. It means only that the parties to the PPCA-CRA Broadcast Agreement determined that total revenue should form the basis for payment for the broadcast right. The use of sound recordings in the simulcast, which is not covered by the PPCA-CRA Broadcast Agreement, remains a use for which payment is required.

120    Having regard to these matters we consider that, if all other things were equal, a payment scheme for simulcasting on a per stream basis would be the only rational, and thus reasonable, option. Such a system enjoys substantial advantages compared to a percentage of revenue payment system from the point of view of both copyright owners and, in reality, the radio stations. As to the latter, the radio stations would be paying for the use of a product on a basis which correlates with their actual use on a direct and transparent basis where, by their own music use choices, they could control the cost of this input. Every station would pay the same amount for the same product. They would do so, moreover, in circumstances where the reality of the current position in the US demonstrates that, subject to the rate per stream being appropriate, the technical and practical burdens imposed by the implementation of this system are nowhere near as difficult as CRA would have it.

121    As we have noted, however, in contrast to the position in the US where there is no payment system in place for the use of the broadcast right, all things are not equal in Australia. Not only is there a long history of payment for the use of sound recordings in broadcasting on a percentage of revenue basis, that basis in all likelihood will continue into the future because it is not possible to identify, directly, numbers of people tuning into any radio station. It follows that one key submission CRA made was valid and of real significance, namely, that if a per stream payment system is imposed then radio stations will be subject to two different systems for two related rights – one for the broadcast right and another for the simulcast right.

122    What we do not accept is CRA’s submission that a hybrid scheme necessarily means that “double reporting is required, in effect, because stations have to do both in order to work out which method is most economic for them”. This will depend on the requirements of the hybrid scheme. If, for example, a hybrid scheme required radio stations to pay on a per stream or percentage of revenue basis whichever is the greater then CRA’s submission would be correct. This, we note, is the formula used in many of the recent webcasting agreements with which we were provided. Those agreements demonstrate that it is by no means impossible or unreasonably burdensome for a single entity to have to implement different systems in order to work out which payment they have to make. By contrast, a hybrid scheme which left it to the radio stations to choose to make a payment on either basis as they see fit would not necessarily involve any radio station in a double reporting system. It would be a matter for each radio station to decide whether there is sufficient justification for it to calculate its liability on a per stream basis.

123    In our view, the advantages of per stream payments for simulcasting are such that it would be reasonable to try to encourage radio stations to implement such a process. Because radio stations already have to comply with a percentage of revenue payment system under the PPCA-CRA Broadcast Agreement the only way in which such encouragement could be given is by an initial discount on the per stream rate that would otherwise apply. When we raised this possibility with the parties, PPCA’s response was that radio stations had already enjoyed free use of the simulcast right for many years and should not be given a further free ride. The difficulty, we recognise, is that the selection of any rate, be it on a per stream or percentage of revenue approach, involves estimation and approximation (as will be apparent from the discussion below), so attempting to identify an effective incentive is nearly impossible. Moreover, every radio station is different. It is not possible to work out the incentive which any discount will present to any individual radio station. As such, a broad-brush approach will be required.

124    While we have some sympathy with PPCA’s position, we consider that the benefits to PPCA’s administration if a number of radio stations adopted a per stream approach would be such as to warrant encouraging the initial required investment of time and money by a short-term (say, first year) discount on the per stream rate that is otherwise considered reasonable.

125    We do not suggest that it would be appropriate for any incentive to be provided by increasing the percentage of revenue rate that would otherwise be applied. Such an approach, in our view, would not be reasonable. The incentive relates to the implementation of the per stream payment scheme and, accordingly, it is reasonable only to reduce that rate below what it would otherwise be for a period.

126    One other observation should be made. CRA made a submission as follows:

Any ulterior motivation, such as PPCA wanting to utilise the simulcast right to compensate it for what it considers to be insufficient royalty income from the broadcast right, is not adequate justification for the Tribunal to consider embarking on such a course [that is, a per stream payment system].

127    We did not discern any ulterior motivation of this kind on PPCA’s part.

128    While PPCA undoubtedly considers the amount paid for the broadcast right to be too little, it was common ground between the parties that the only value in issue in this case is that of the simulcast right. It goes without saying that the value of the simulcast right cannot be inflated to make up for some perceived deficiency in the value of the broadcast right. PPCA did not suggest this should occur and we accept it cannot occur.

4.    RATE SETTING – GENERAL

4.1    The simulcast right

129    The right is to use sound recordings in a radio simulcast subject only to a capacity to replace advertising.

130    Accordingly, the right depends on there being a radio broadcast which uses sound recordings which is transmitted not only via the radio frequency spectrum but also over the internet. As CRA submitted the right does not include:

(a)    webcasting (a licence for this is an additional requirement for ARN, for example, for its customised radio services provided through iHeartRadio);

(b)    podcasting (podcasts are downloadable episodes featuring "all the best bits" from radio programmes, including presenter talk-back, celebrity interviews and competitions and generally contain no music and no ads);

(c)    reproductions; nor

(d)    any use of sound recordings in other website activities.

131    One issue that may be disposed of immediately is CRA’s characterisation of the simulcast right as ancillary to the broadcast and thus not a separate economic activity in its own right.

132    There is no dispute between the parties that the simulcast cannot occur without the broadcast. The simulcast is simply the broadcast as transmitted via the internet (albeit with a capacity to which PPCA agrees for a radio station to substitute different advertisements in the simulcast). The simulcast is thus ancillary to and dependent upon the broadcast. But in the present matter it is a scheme for the licensing of the simulcast which must be determined. The activity of simulcasting is thus the necessary focus. Whether the activity is ancillary or not it involves a right which must be valued. Labels such as “ancillary” or phrases such as “not a separate economic activity” are not particularly helpful in determining the value of the right.

133    Consistent with this view, we do not agree with CRA’s submission that:

At the heart of the dispute between the parties is the characterisation of the radio stations’ delivery of radio programmes. CRA says that a station is engaged in a single economic activity. PPCA contends that simulcasting involves the separate exercise of a copyright with no relation to the broadcast, and creates a separate revenue stream for the radio station. This gives rise to competing approaches to the form of the royalty regime to apply to the exercise of the so-called simulcast right in relation to radio programmes.

134    Rather, in our view, at the heart of the dispute is a substantive issue about the value of the simulcast right and how radio stations should pay for that value.

135    We also do not agree with CRA’s submission that “the task of the Tribunal is to assess the additional economic benefit delivered to the radio stations by the exercise of the ‘simulcast right’ (if any)” if, by this, CRA considers value to mean nothing more than any additional or incremental profit that is directly attributable to simulcasting (for which CRA cited in support the 2MMM case at 25). As explained below, simulcasting performs functions for radio stations currently and has the potential to perform further functions. This means that simulcasting has a value over and above the revenue which can be directly attributed to that activity.

136    Further, the 2MMM case at 25 ultimately adopts a notional bargain approach. This approach involves consideration of the value which the parties would have agreed in all of the circumstances in an arm’s length negotiation assuming only a willingness to conclude an agreement. For reasons expanded upon below we consider that an attempt in any notional bargain to confine the value of simulcasting to the revenue which can be directly attributed to that activity would not succeed. The concluded bargain would have to be based on the full range of benefits which simulcasting provides to radio stations.

137    CRA also made the following submission which we find difficult to understand:

fundamental to PPCA’s rate structure is the proposition, propounded by it for the first time in its history, that sound recordings embedded in radio programmes have an inherent value unrelated to the revenue that is earned from the radio programmes.

138    The submission appears to suggest that a sound recording has no inherent value and derives its value only from the revenue that a radio station can earn. This confuses value with the method which thus far has been chosen as a proxy for value in the context of use of sound recordings by radio stations. Radio stations, however, are but one user of sound recordings. Other users, such as webcasters which provide music streaming services, do not necessarily pay on a percentage of revenue basis. The percentage of revenue basis for payment is a proxy for the value of the simulcast right, but not the only possible proxy. It follows that simulcasting does have an inherent value separate from the revenue that a radio station can earn from its programs. The issue in the present matter is identifying the value of the right and the method by which radio stations should pay for the use of that right.

4.2    Value of the simulcast right to radio stations

139    PPCA characterised the benefits that radio stations currently obtain and could obtain if they so choose from simulcasting as substantial. CRA characterised them as marginal, relying in particular on the fact that many stations ceased to simulcast when the interim scheme was imposed and had not suffered any apparent loss of revenue.

140    The first point to make is that the cessation of simulcasting by regional radio stations due to fears about the interim scheme and, most likely in our view, to bring political pressure to bear in respect of the need for a simulcasting licence at all is not probative of anything other than the existence of fear of the unknown and a belief in the effectiveness of collective action. It does not prove that those stations considered simulcasting to be unimportant or insignificant. In any event, beliefs about the importance or lack of it of simulcasting have to be rational and justifiable before they are entitled to material weight in the assessment of the notional bargaining process. This is why the concept of two parties, with all their strengths and weaknesses and an assumed willingness to conclude a bargain, is critical to the concept of a notional bargain. In the process of the assumed bargain, misconceptions, ill-founded beliefs and exaggerated concerns are unlikely to survive or to have any real impact on the bargain ultimately reached.

141    The fact that, when simulcasting became available, most radio stations chose to use the new technology suggests that radio stations saw the opportunity, which involves a relatively low start-up cost, as one they should pursue. This is not surprising. Radio stations live or die by audience numbers. It is imperative they gain and retain listeners. Simulcasting provides another method of gaining and retaining listeners. Given that the costs of adopting the new technology were low, and the radio stations were paying no additional fees to enable simulcasting, it would have been irrational for a station not to make a simulcasting service available to its listeners. Characterising simulcasting as marginal or unimportant misses the critical fact – for a radio station every method of gaining and retaining a listener is important. Such a characterisation provides no real assistance in valuing the simulcast right. That value is not determined by labels such as important or marginal – it is determined by the price for the use of the sound recordings at which the copyright owners are willing to sell and the price for that use the radio stations are willing to pay.

142    The second point to make is this – simulcasting already accounts for about 10%, on average, of listeners to radio stations (the proportion being greater than this in metropolitan areas and lesser in regional areas). Given that radio stations make advertising revenue (their overwhelmingly dominant revenue source) based on audience share, it is not the case that the simulcasting right has a zero or nominal value as Dr Epstein and Mr Samuel maintained. CRA’s submissions correctly noted in this regard that simulcasting “accounts for approximately 10% of the radio audience and therefore, in effect, 10% of the revenue derived from programmes simulcast, since the simulcast contains no additional or different advertising from the broadcast”. On this basis, to approach the simulcast right based on a preconceived belief that, because the activity is ancillary to broadcasting, it must have no or nominal value (as Dr Epstein and Mr Samuel appear to have done), is erroneous.

143    The fact of the rapid switch-off by regional radio stations in response to the interim scheme is not a reliable basis for inferring anything about the value of simulcasting to radio stations. It might equally be said that the switch-off discloses that radio stations believe simulcasting is important because many of them (or at least the regional stations) were prepared to cease the activity, at least in part in our view, to achieve the policy objective of ensuring that simulcasting was covered by the PPCA-CRA Broadcast Agreement. A more reliable indicator of a present value of the simulcast right greater than merely nominal value is the average proportion of listeners who listen via the simulcast.

144    We accept that we cannot predict the future for listening via the simulcast. But the submission advanced by CRA that the most recent survey figures indicate a downward trend is untenable. It is difficult to see the trend heading in any other direction than upwards although by what percentage each year is unknown. For present purposes we are prepared to accept the overall average of something over 10%. In the radio business 10% of the audience is substantial. To understand the significance of this 10% figure, the evidence disclosed that a one percent share of audience in the Sydney broadcast market equates to approximately $5 million in advertising sales.

145    The third point to make is this – simulcasting carries with it opportunities for exploitation that are not present for the broadcast. As already noted, simulcasting enables a radio program to provide its listeners with access in any geographic location. It enables advertisements to be targeted to the audience. These capacities will never form part of the broadcast because of the nature of the technology. But they are available for the simulcast. This difference in capacity indicates that, even if (as Ms Dick’s evidence indicates), advertisers in Australia are currently largely indifferent to the potential advertising benefits of the simulcast, this is not likely to remain the case in the foreseeable future.

146    The fourth point to make is this – whether they acknowledge it or not radio stations, particularly those that play a significant component of contemporary music, are in competition for listeners with online music service providers. While we accept that the products offered by radio are different from online music providers in the way CRA described in that a radio station provides a package of entertainment, radio stations are nevertheless competing (at least in the broadest sense of that concept) for listening hours with online music service providers.

147    We therefore accept the submission for PPCA that, in circumstances where the current simulcast audience comprises about 10% on average of total radio listening, if radio stations ceased simulcasting over the long-term they would be likely to lose some of their audience to online music service providers. We do not accept that the 10% average simulcast audience is wholly cannibalised from the overall radio audience. But for simulcasting, over the longer term, we consider a material loss of overall audience share by radio stations to online music service providers is likely. Simulcasting thus also has this additional defensive value of enabling audience share to be maintained (even if not increased), into the future.

148    The fifth point in this regard is that, despite CRA’s claims to the contrary, there is potential for radio stations to earn money from their simulcasting activity. While we accept that this potential has not been exploited as yet in Australia (as well as the UK, Canada and New Zealand) the position in the US is more developed and it seems inevitable that Australia (and the other countries) will follow suit. As PPCA submitted, internet radio has the advantage that advertisements can be “targeted, interactive, and can include a visual element, and that the audience can be reached via work computers, smartphones, tablets and social media”. We find it difficult to imagine that these capacities will not be exploited in the foreseeable future.

149    Much effort was expended by the parties on the issue of the relationship between the websites of radio stations and the simulcast. We accept that the website is one thing and the simulcast another. We accept also that, unlike the position with the simulcast, “websites of major Australian radio stations already demonstrate a strategic attempt to capitalise on the value of an online audience”. But the fact is that listeners access the simulcast via a function on the websites of radio stations. The desire to listen to the simulcast thus draws the listener to the website. Conversely, being on the website is likely to draw the listener to the simulcast. Despite CRA’s assertions to the contrary, the same cannot be said of the broadcast. We therefore accept Dr Coker’s evidence that:

The possibilities for interactivity via the simulcast or website are much more intuitive and fluid that those available historically and, critically, are available to all listeners of the simulcast. There are multiple possibilities for engagement available, for example, listeners can vote on the popularity of songs while listening, can view video of radio announcers in the studio, comment on music and social features, enter competitions and link to social media. All these possibilities for interactive engagement are more seamless and immediate because they all occur via internet connectivity. It is well-established that increased interactivity is a strong predictor of listener loyalty and as such is likely to be an important benefit both to consumers and radio stations seeking to maintain their ratings…

150    In this context, it is difficult to accept that radio stations in Australia will not seek to monetise their simulcasting activities to a greater extent in the near future than is presently the case. This potential has a present value which must be recognised.

151    Insofar as it was a common theme of CRA and the representatives of radio stations that the simulcast right had no value to them at all or at most only a marginal value, we disagree for the reasons given above. We do not suggest that those giving evidence were doing other than giving their honest opinions. We consider however that those opinions do not accurately reflect the reality of the situation and are likely to have been affected at least by unconscious self-interest.

152    This does not mean that music necessarily has the same value to radio stations as it does to online music service providers or that radio stations are in the same bargaining position as those providers. Music is the only thing that online music providers offer to listeners. Many online music providers are unproven start up enterprises. They negotiate as individual enterprises with the major record companies. None of these propositions apply to radio stations. Commercial radio is an established industry in Australia capable of unified activity and bringing to bear collective bargaining strategies. Radio stations do not necessarily depend on music. While some stations use music a great deal, most also provide other forms of entertainment, such as talk by popular radio personalities, news, sports and weather. This is so for nearly all radio stations. While PPCA was able to identify a station which had a music use percentage of over 97%, the average music use percentage is around 61%. While this is significant it means radio stations offer an entertainment package involving things other than music.

153    Equally, while some radio stations use their dominant music style as an important part of their branding, others do not. PPCA rightly pointed out that some radio stations describe their formats using terms such as “contemporary hit radio”, “easy listening”, “mainstream rock”, “chill out”, and “hot adult contemporary”, which indicates that music is an important part of the package being offered. But this does not change the fact that generally speaking radio stations offer more than music, a fact which sets them apart from online music providers.

154    The evidence and submissions to the effect that radio station presenters are more important than music or music is equally or more important than radio station presenters involve one of those imponderable issues which need not, and probably cannot, be resolved. Radio stations use music and presenters. The Tribunal’s task is to assess the reasonableness of schemes for the use of one (the sound recording) in the simulcast, and not the other. As PPCA put it, the fact that some presenters are very important to the success of a radio station is of little relevance to the reasonableness of the schemes.

155    For these reasons we consider that the bargaining position of radio stations relative to PPCA is different from that of online music providers. Radio stations in Australia are in a materially better bargaining positon than those providers. Much as PPCA and copyright owners would maintain that there should be a level playing field so that all who use sound recordings should pay the same for the use, this does not reflect market reality. Producers do not charge all customers the same amount for the same product. Some buyers are capable of negotiating a better rate because they have a better bargaining position than other buyers and it is still worth the producers while to sell at that rate. In any notional bargain between PPCA and radio stations the factors to which we have referred would be brought to bear and result in the radio stations being able to negotiate a better rate for their use of music than that achieved by online music providers.

156    In summary, we do not accept that a reasonable scheme is one which attributes a nominal value only to the simulcast right. The simulcast right has material value even when considered from the perspective only of radio stations. However, the value to radio stations is not necessarily the same as the value to webcasters. It also has value, of course, when considered from the perspective of copyright owners, the issue to which we now turn.

4.3    The simulcast right – its value to copyright owners

157    The evidence disclosed that the sound recording industry has been subject to rapid and sweeping changes over the past 15 years, mostly wrought by the development of digital technology. The entire nature of that industry has changed dramatically over those years. It appears that the revenue basis of the industry has moved from an “acquisition” to an “access” model. Before digital technology people acquired music by buying a record, tape, compact disc or (in the initial digital technology phase) by download. Even the phase of the downloading of music (from iTunes and the like) proved short-lived in this rapidly changing environment. Currently, people largely access rather than acquire music. As such, the record companies have had no choice but to respond to the changing environment. If they are to remain viable, they have to ensure a reasonable return on the right of access to the product, the music.

158    The product (sound recordings), moreover, is by no means risk free for a record company. As was disclosed in the evidence there is substantial effort and risk involved for a record company in producing a sound recording. The evidence indicated that most new artists do not become commercially successful and that, even if a new artist does enjoy initial commercial success, such success is often short-lived and does not carry through to subsequent recordings. Mr Massarsky, for example, characterised the effort, risk and reward relationship in respect of sound recordings as very different from that in respect of song-writing, an opinion which we accept. The pressures on record companies created by that effort, risk and reward relationship are significant. The fact that they are dealing with these issues in a rapidly changing environment makes good the proposition that they must ensure a reasonable return on the use of sound recordings in order to maintain viability.

159    That this is so is supported by PPCA’s submission, which accurately reflected the evidence, as follows:

Ms Don described the collapse of record companies’ traditional revenue streams that accompanied the digitisation of music distribution in the early 2000s, and the increasing importance of licensing revenue in their reformed business models. In the early 2000s, record companies’ main source of revenue was physical music sales – for example, in 2003, 92% of Universal’s revenue was derived from physical sales. ARIA figures show that the value of physical sales to the Australian record industry dropped from $606,970,000 in 2004 to $130,067,254 in 2014, a decline of almost 80%. Ms Don described record companies’ efforts to respond to that precipitous decline by developing new sources of revenue and ensuring that value is derived from all such sources…

While these efforts have had some success, ARIA figures show that the total value of the Australian recorded music industry has nonetheless declined from $609,970,000 in 2004 to $317,790,300 in 2014, in real terms, a decline well in excess of 50%...

The decline in the value of music sales is attributable to the digitisation of music distribution, and the shift from a “music ownership” to a “music access” model. Ms Don explained that a key part of record companies’ response to that change has been to seek to maximise revenue from licensing, especially from streaming services, such as Spotify and Pandora, which are important distributors of sound recordings through the music access model. While digital revenues remain flat, because sales of digital downloads (such as from Apple iTunes) are not growing, revenue from streaming services such as Spotify and Pandora is growing. Ms Don explained that streaming services provide “a critically important core – not incidental – revenue stream in [Universal’s] more diversified business model.”

…

Mr Massarsky’s oral evidence reinforced the importance to record companies of royalties following the shift from a music ownership to a music access model:

Today everything counts. Everything is now under inspection … the record industry now, that has become reliant upon subscription models and digital licensing models and what we still call access models …

160    We accept, moreover, Ms Don’s evidence that:

… it’s my view that all operators who are operating in this digital space should be paying a reasonable amount for the use of the content on some sort of, you know, equal level, in broad terms …

So if a commercial radio station came to us and sought a license directly, then we would approach that negotiation in the same way as if we were negotiating with, say, Samsung Milk or Pandora.

161    We accept that this would be the negotiating position which a record company such as Universal would adopt if an individual radio station sought a licence for use of sound recordings. For the reasons given, however, we do not believe that this scenario represents the notional bargain which we are attempting to identify. This is not a case in which the negotiation is between an individual radio station and a record company. We are attempting to determine a scheme as between all copyright owners which PPCA represents and all radio stations which wish to simulcast. The notional bargain is thus collective. Moreover, from the point of view of the radio stations, they would rightly see themselves as enjoying a stronger bargaining position than an online music service provider. It is for these reasons that we do not accept that it is likely that the bargain that would be reached in the present case would result in radio stations paying the same amount for use of sound recordings as online music service providers.

162    This is so despite the fact that we accept Mr Massarsky’s opinions that, currently, radio stations do not have the same critical role that they used to in introducing new music. Given the many sources now available for new artists to be introduced, this opinion is inherently persuasive. For this reason we accept PPCA’s submission that the relative importance of radio as a promotional outlet for sound recordings has decreased with the rise of digital music distribution. It follows that to the extent that CRA’s case involved the suggestion that there was effectively a quid pro quo between radio stations and copyright owners, which supported its case that the simulcasting right had no or merely a nominal value, we do not accept it. We consider that some radio stations provide new music with a promotional opportunity but the importance of that promotional opportunity has substantially reduced over recent years given the large number of other opportunities that now exist for the public to become aware of new artists. Accordingly, the promotional opportunity presented by airplay of new music on the radio would be a minor or immaterial factor in any notional bargain between PPCA and CRA.

163    We have difficulty with Dr Epstein’s opinion that requiring radio stations to pay anything more than a nominal amount for simulcasting is economically inappropriate because it will limit output. The only “output” Dr Epstein appeared to have had in mind in this regard is simulcasting by radio stations. Dr Epstein considered that the fact that many regional stations ceased simulcasting in response to the Tribunal’s approval of the interim scheme demonstrated that the economic inefficiency of output limitation had already occurred, which was undesirable. Three comments in response are appropriate.

164    First, and as discussed, the notion that the regional radio stations were acting as pure economic rationalists when they decided to cease simulcasting is inconsistent with the evidence. They were acting because of fear of the unknown and, it seems likely, because they believed ceasing simulcasting would assist a political campaign to pressure government to change the law so that the broadcast right would also cover simulcasting.

165    Second, Dr Epstein appears not to have given any real weight to another relevant output – being the production of sound recordings.

166    Third, radio, either by broadcast or simulcast, is only one medium by which sound recordings can be communicated to the public. It is not the only, or necessarily the most important, medium.

167    Otherwise, insofar as PPCA suggested that the value of the simulcast right should be greater than it might otherwise be to account for the impacts of digital piracy on the sound recording industry, we disagree. Digital piracy is relevant in that it is a fact with which record companies (and PPCA as their representative) must contend. In this sense, it is relevant to the price that the record company (or PPCA) might wish to achieve for the right to use a sound recording. But it cannot dictate the price reached in a notional bargain with the other party, the radio stations, which are not responsible for digital piracy and would not see their role as subsidising record companies to make up for losses which digital piracy causes. In our view, it would not be reasonable to adjust the price that would be achieved in the notional bargain to subsidise record companies for the effects of digital piracy.

168    For these reasons, we consider that PPCA is right to insist that the simulcast right has more than a merely nominal value and that CRA’s attempts to place a nominal value on that right are unrealistic. PPCA is not right, however, to insist that radio stations are in the same position as webcasters and thus would pay the same amount to use sound recordings as webcasters.

4.4    Using commercial benchmarks

169    We accept the opinion of Dr Eisenach that if relevant commercial benchmarks for the value of a right are available, then it is preferable to use those benchmarks to value the right rather than to use synthetic methods such as economic modelling. Economic modelling depends on the input assumptions. Small adjustments in inputs can lead to vastly different outputs. Assumptions are also often based on subjective preferences. In the present case, presumably for these reasons, none of the experts approached the issue of the value of the simulcast right on the basis of economic modelling. All focused on what they considered to be commercial benchmarks.

170    We have referred above to “relevant” commercial benchmarks. This is because, in common with Dr Eisenach, we see nothing useful in becoming bogged down in the terminology of comparable or non-comparable bargains. Like Dr Eisenach we consider the comparability of a bargain to be a relative concept. The relativities are best conceived of as a scale sliding from directly comparable at one end to non-comparable at the other. Even a bargain which is non-comparable may disclose something helpful to the attempt to identify the notional bargain between the parties. A non-comparable bargain is not necessarily irrelevant, although it might not form a sound basis for adjustment to equate to the required notional bargain. That said, it is unlikely that any bargain will be directly comparable and require no adjustment for the purpose of the notional bargain.

5.    RATE SETTING

5.1    Overview of PPCA’s case

171    PPCA submitted that there were two groups of comparable bargains for a per stream payment system, being:

(a)    the agreement entered into in the US in 2009 between the National Association of Broadcasters (NAB) and SoundExchange (the NAB-SX Agreement) licensing the use of sound recordings in simulcasts of radio broadcasts; and

(b)    recent webcasting agreements involving the major record companies Universal, Sony and Warner, under which the use of sound recordings in radio-like online music streaming services is licensed in Australia and elsewhere.

172    PPCA also submitted that other bargains were not comparable but were useful as they provided a benchmark of the total fees payable against which the reasonableness of the fees payable under the scheme for simulcasting could be assessed. Those agreements were:

(a)    the “Industry Broadcast Agreement” between PPCA and CRA entered into on 16 June 2000 (the PPCA-CRA Broadcast Agreement);

(b)    the musical works licences between APRA and commercial radio stations (referred to collectively in the evidence as the APRA-CRA Agreement);

(c)    the older, experimental and non-precedential webcasting agreements between PPCA and various online music streaming services; and

(d)    other overseas régimes, including those in the UK, New Zealand and Canada.

173    PPCA adopted the approach of Dr Eisenach that the bargains it described as comparable captured the value of both the existing revenues attributable to simulcasting and all of the other benefits which simulcasting provides (as described above), including the present value of its potential. As Dr Eisenach put it:

I accept as a lower band, this attribution of 10 per cent of advertising – 10 per cent of the audience as a rough justice perhaps estimate of the immediate value of the advertising attributable to simulcasting. I don’t think it’s representative of the total value of the simulcasting to the radio broadcasters for the reasons that I described earlier…

It’s just a much bigger picture in terms of the value that the future – the discount at present – net present value of future earnings associated with the ability to – to capitalise on that business opportunity.

174    Dr Eisenach described the NAB-SX Agreement in these terms:

The best way, in general, in economics and valuing, a good idea is to look at what a willing buyer and a willing seller agree upon in the marketplace.

And in this case the conclusion that I have come to is that in the US the NAB/Sound Exchange agreement is such a bargain, a gift in a way. And the cases that I’ve been involved in and that I’ve studied, it is rare to find an agreement which is as closely analogous to the agreement that one is trying to value as the NAB/Sound Exchange agreement is – here we have virtually, practically identical rights, we have analogous parties and, as I will discuss, we have markets which are very close in the relevant characteristics.

And the conclusion – to summarise briefly my conclusions – and I will talk a little bit more about my methodology – to summarise briefly my conclusions, the US Sound Exchange/NAB agreement is very much a comparable bargain and the rates included in it, as I show in my testimony, are rates which are appropriate in the circumstances, I believe, for the PPCA scheme.

175    He described the APRA-CRA Agreement (relating to the right to use musical works) as useful, observing as follows:

The APRA agreement is an agreement that involves multiple sets of rates including the simulcast rate as well as the broadcast rate. So as one thinks about that, the ability to engage in broadcasting and simulcasting that is enabled, permitted by those two rates together is equivalent to the permission – the economic activity is equivalent to two different rates in this circumstances, the broadcast rate and the simulcast rate.

So my conclusion with respect to the APRA/CRA agreement is that it is reasonable to look at that as an indication of the willingness to pay of broadcasters for that combined set of rates. I emphasise “willingness to pay”; we have a circumstance in which those two sets of rates which enable broadcasting and simulcasting or necessary for broadcasting and simulcasting, we have an instance where we know what, in fact, is being paid for those rights. So I think it’s relevant to look at the total amount as I’m sure we will get into. The difficulty in disaggregating in using the CRA/APRA agreement as a comparable bargain in order to establish the value of the simulcast rate is that, from an economic perspective, the simulcast rate and the broadcast rate are perfect complements in the production of simulcasts.

…

…the relative value of those two perfect complements is indeterminate from an economic perspective. It can’t reasonably be disaggregated in any meaningful way. So while I think the APRA total amount is a reasonable benchmark in terms of thinking about the reasonableness of the total licence fees to be paid, I don’t think that the APRA agreement can be used as Mr Samuel and Mr – Mr Samuel in particular has to arrive at a simulcast rate.

176    Mr Massarsky largely shared Dr Eisenach’s views. He said:

First of all, with the SoundExchange right, we are talking about the right use type, a sound recording. A sound recording operates very differently than APRAs use of a representation of a work or a composition. They are two very different rights and they have very different returns on investment and financial models. So we have the right method of measurement when we look at the SoundExchange right. Secondly, we don’t have to worry about unbundling. It’s already unbundled for you. It’s the simulcast right that pertains to commercial radio stations. It’s a very effective way to look at a comparable bargain.

Third, when we really study the difference between the broadcast right between CRA and PPCA and the proposed simulcast right, what really jumps out of the page, a real big differentiator that we have to deal with in the next three days, is what the value of the unprotected sound recording rights? Those unprotected rights have no value within the broadcast scheme, but because of the nature of copyright open up great opportunities for the unprotected side, and most, if not all, of the unprotected works are, in fact, US works.

So here we have this excellent comparable, a sound recording, unbundled, in its original form, as a simulcast, and we’re measuring as we need to do in Australia, the unprotected work’s function which primarily are the US music…

5.2    Overview of CRA’s case

177    CRA submitted that its approach was reasonable and was supported by the evidence of Dr Epstein and Mr Samuel. It relied on Dr Epstein’s opinions which it summarised as follows:

(a)    the simulcast audience for commercial radio in Australia is virtually indistinguishable economically from the broadcast audience. CRA members treat earnings from advertisements embedded in the radio programming as a consolidated amount, regardless of the platform of listening, and have therefore effectively been paying a royalty to PPCA for simulcasts since the inception of simulcasting;

(b)    the current broadcast royalty of 0.4% in the Industry Broadcast Agreement, which has been in effect for more than 13 years, is the most comparable benchmark and the appropriate starting point for analysing a rate for simulcasts. The so-called “1% cap” is not a binding constraint;

(c)    a percentage of revenue scheme for simulcasting is consistent with the long-standing scheme used for the broadcast right in Australia, is efficient, simple to calculate and administer, and ties the royalty to the economic value of the rights. It is also the method used for the APRA|AMCOS Agreements with CRA (which bundles together both the broadcast and simulcast rights, as well as other rights); and

(d)    the per-stream scheme proposed by PPCA represents an enormous increase compared to the existing broadcast royalty, lacks a connection to the value of the rights in Australia (being the revenue earned from simulcasts), and is economically inefficient, both through restricting the output of simulcasts and being more costly to administer. He considers that an output reduction from PPCA's proposed scheme is not merely a potential; it has in fact occurred, with a number of stations ceasing to simulcast due to the risk of back pay based on PPCA's proposed rates.

178    CRA explained Dr Epstein’s methodology for calculating an appropriate percentage of revenue payment rate in these terms:

88.    Dr Epstein concludes the appropriate methodology for determining a royalty for simulcasting is as follows:

(a)    starting with the broadcast rate of 0.4%, take account of royalties for currently unprotected sound recordings that are not captured in the Industry Broadcast Agreement (by grossing up the broadcast royalty);

(b)    take account of the fact that stations already effectively pay a royalty under the Industry Broadcast Agreement for protected sound recordings based on revenue attributed to advertisements embedded in the simulcast stream; and

(c)    take account of any direct simulcast advertising revenue that results when a listener starts the simulcast stream (eg, pre-roll advertising) that is not already included in CRA members' gross revenue. He does not consider that other website revenue should be included, as station websites typically contain a great deal of content that is accessed separately from the simulcast stream.

89.    The application of this methodology results in a royalty rate of 0.036% of CRA members' gross revenue, plus a rate of 0.741% of total direct simulcasting advertising revenue. These rates assume a protected/unprotected works split of 54/46 and a simulcast audience percentage of 10.7%, although different inputs can be used. In his second report, in order to make more explicit how the appropriate royalty varies as the simulcast listener share changes, Dr Epstein repeats the methodology for his royalty expression, but with the simulcast listener share as a separate element in the analysis (ie, the formula can be more clearly expressed as a royalty rate of 0.341% of gross revenue times the simulcast listener share, plus 0.741% of total direct simulcasting advertising revenue).

179    CRA provided a table which demonstrates the effects of different inputs on the calculation of percentage of revenue rates in accordance with Dr Epstein’s proposed formula, as follows:

180    The protected/unprotected proportions refer to the fact that the broadcast right applies only to sound recordings where there is equivalent copyright protection from the copyright country of origin. For example, US sound recordings are unprotected because there is no right equivalent to the broadcast right in the US.

181    Insofar as Mr Samuel is concerned, CRA submitted as follows:

93.    Mr Samuel considered the proposed royalty structures in the following manner:

(a)    he identified the market approach (by reference to comparable bargains) and the income approach (by reference to incremental cash flows derived by the licensee) as the appropriate methodologies for considering a reasonable royalty rate;

(b)    he considered the potential royalty rate structures for the proposed licence (percentage of revenue, price per unit, fixed amount and minimum amount); and

(c)    he concluded that a percentage of revenue structure is the most appropriate and reasonable structure in the circumstances of the Australian market, particularly given that simulcasting has not generated any new (or any materially new) incremental revenue or profit for radio stations and, as a consequence, the appropriate royalty for both protected and unprotected sound recordings would be nil or nominal.

182    CRA explained Mr Samuel’s methodology in these terms:

94.    Mr Samuel considers the appropriate methodology for determining a royalty for simulcasting is as follows:

(a)    adjust the royalty under the Industry Broadcast Agreement to reflect the relative increase for simulcasting from:

(i)    the licences between APRA and the commercial radio stations entered into in or around December 1999 (the Previous APRA Agreement);

(ii)    the licences between APRA, AMCOS and the commercial radio stations entered into in or around April 2010 (the Current APRA Agreement),

(collectively, the APRA|AMCOS Agreements); and

(b)    in making this adjustment, allow for the fact that the Industry Broadcast Agreement does not cover unprotected sound recordings (an issue which does not exist in the APRA|AMCOS Agreements or the proposed simulcast licence).

95.    The application of this methodology results in a royalty rate of 0.065% of gross revenue.

183    It will be apparent from this that CRA’s methods attribute to the simulcast right a present value of between 0.0341% and 0.065% of gross revenue.

184    Mr Samuel also proposed an alternative methodology which did not use the PPCA-CRA Broadcast Agreement as the starting point but, rather, the APRA-CRA Agreement. This method is discussed separately below.

5.3    The PPCA-CRA Broadcast Agreement

185    The PPCA-CRA Broadcast Agreement was entered into in June 2000 by PPCA and CRA’s predecessor, The Federation of Australian Broadcasters Limited (FARB), and was deemed to commence on 1 July 1999. Under the agreement, CRA is bound to collect the gross licence fee from its members and to pay that amount to PPCA, in consideration for which PPCA grants to those members a licence to use protected sound recordings in broadcasts.

186    Schedule 1 to the PPCA-CRA Broadcast Agreement identifies the gross licence fees payable as follows:

187    We explain below our reasons for concluding that the PPCA-CRA Broadcast Agreement does not provide a reliable basis for drawing any conclusions about the value of the simulcast right.

188    Before doing so we note that CRA submitted that this case “does not call for a determination of the reasonableness of the broadcast rate and/or the reasonableness of the 1% statutory cap” as “PPCA has chosen not to submit a scheme for broadcasting to the Tribunal for determination”. The problem with this submission is that it is CRA, not PPCA, which seeks to use the broadcast rate of 0.4% as the starting point for its proposed schemes. In so doing it is CRA which has necessarily raised the issue whether the broadcast rate is an appropriate starting point. It is impossible to determine the reasonableness of the schemes which CRA has proposed on any other basis.

189    As noted, s 152(8) of the Copyright Act provides that the Tribunal may not require any broadcaster to pay more than 1% of its gross revenue for a broadcast licence. To the extent that CRA submitted that this was a constraint on the Tribunal only and did not prevent PPCA negotiating a higher rate with radio stations, only one response is required. No radio station would agree to pay more than 1% in circumstances where it knows that the circuit-breaker for any deadlock in negotiations, the Tribunal, is confined by the 1% cap.

190    Before the PPCA-CRA Broadcast Agreement was negotiated the 2MMM case (also referred to as the WEA case) had been run in the Tribunal as a test case. The background to this occurring is identified in a discussion paper released by the Attorney-General on 1 February 2005 (Review of One Per Cent Cap on Licence Fees Paid to Copyright Owners for Playing Sound Recordings on the Radio). The discussion paper records the following:

29. The broadcast licence cap applies only to the broadcast right in sound recordings. There is no similar limit placed on the Copyright Tribunal in determining the licence rates that broadcasters must pay to APRA for broadcasting works embodied in a sound recording – only that the amount be ‘equitable remuneration’ to the copyright owner.

…

31. The Copyright Act 1968 commenced on 1 May 1969. Shortly before that date, the PPCA was established by the major record companies as their collecting agent and to issue licences for the broadcast and public performance of their sound recordings.

32. A request made by PPCA to the commercial radio stations for royalty payments initiated an open dispute. The broadcasters refused to pay royalties or give an undertaking to pay royalties under the s 109 broadcasting licence. From May to November 1970 commercial radio stations imposed a ban on playing ‘protected’ sound recordings, i.e., recordings which attracted broadcast licence fees (see para 60 below). During the ban stations played only records without a broadcasting right, particularly United States records, and local records whose manufacturers did not insist on royalties. The ban ended when PPCA agreed to licence the commercial broadcasters in return for advertising credits (three 30-second announcements per week on each station in prime time).

33. In September 1980 the major record companies made applications to the Copyright Tribunal for orders determining the amounts payable by several FM broadcasting stations under s 152 of the Copyright Act. The commencement of commercial FM radio broadcasts created new pressures on voluntary licence arrangements as the record companies considered that commercial FM radio stations were making far greater use of recorded music than the AM stations. In the negotiations, PPCA offered to licence FM broadcasting for an annual fee of one per cent of a station’s gross earnings while FM broadcasters insisted on the same free airplay terms provided for in the license agreement with the AM stations. PPCA believed free airplay would not provide adequate compensation for the higher use.

…

36. After a lengthy hearing the Copyright Tribunal fixed the amount payable by 2MMM as 0.45 per cent of that station’s gross earnings (WEA Records v Stereo FM (1983) 48 ALR 91). A relevant factor for the Tribunal in determining the contribution made by protected recordings to the earning capacity of 2MMM was the proportion of broadcasting time as a whole occupied by the playing of protected recordings less advertisements. This was 48 per cent in 1981 and 50 per cent in 1982. The Tribunal also took into account and balanced a number of competing factors such as the benefits accruing to record companies through the increased sales of recordings as a result of their exposure through airplay weighed against the then popularity of home taping from the radio.

37. The one per cent cap was a consideration in the WEA case. It may be argued that the one per cent cap was a real constraint on the record companies’ case in that it set a maximum amount that they might claim from a radio broadcaster in any circumstances. On the other hand, it is not certain that the Tribunal would have determined that 2MMM should pay a rate any higher than 0.45 per cent if the one per cent cap had not existed.

38. Neither the record companies nor the commercial broadcasters have made any further application under s 152 for the Copyright Tribunal to determine the royalties with respect to broadcasting sound recordings.

191    In the 2MMM case the Tribunal repeatedly referred to the 1% cap. The Tribunal’s reasoning included the following statements, amongst others:

    “The prohibition imposed by s 152(11) against the Tribunal requiring a commercial broadcaster to pay an amount exceeding 1 per cent of the gross earnings of the broadcaster during the relevant period to which the order applies fixes a ceiling which the Tribunal cannot exceed. The sub-section does not require the Tribunal to approach the task of determining the relevant amount payable by reference to a formula expressed as a percentage of the broadcaster's gross earnings not exceeding 1 per cent, but there are obvious sound practical reasons why the assessment of the relevant amount is best expressed as a percentage of gross earnings” (at 1 IPR 54).

    “In our opinion the amount payable by 2MMM to the owners of copyright in sound recordings in respect of the period commencing 1 October 1980 and ending on 30 June 1983 is best expressed as a percentage of gross earnings attributable to the use of protected recordings. We have considered all relevant matters for the purpose of s 152(7) and have come to the conclusion that the amount payable should be 0.45 per cent of the gross earnings of 2MMM” (at 1 IPR 55).

192    This conclusion was reached in circumstances where the Tribunal had found that 54% of the sound recordings played on 2MMM were protected sound recordings.

193    The Tribunal’s approach provides a rational basis for the position which PPCA adopted thereafter. According to Ms Small’s evidence, PPCA understood that the Tribunal had recognised that it was bound by the 1% cap and had determined the appropriate rate by reference to the percentage of protected sound recordings used in the broadcasting activity. PPCA also understood, reasonably in our view, that as the best it could ever do before the Tribunal in a referral relating to the broadcasting right was a percentage of 1% as related to the use of protected sound recordings, its time and energy were better focused on attempting to lobby government for the removal of the 1% cap from the Copyright Act. Given the time, cost and effort involved in proceedings before this Tribunal, including in the 2MMM case itself, PPCA’s approach makes commercial sense. Suffice to say that the material establishes that PPCA and CRA each lobbied hard for their respective positions and, as we understand it, continue to do so.

194    In these circumstances, the notion that the rates negotiated in the PPCA-CRA Broadcast Agreement represent freely determined market rates, unaffected by any constraint associated with the 1% cap, is beyond the bounds of possibility. It became apparent that neither Dr Epstein nor Mr Samuel was aware of all circumstances relevant to the resolution of this issue.

195    First, as to the fact that the broadcast rate had at one time been as low as 0.2%, and was phased in from 0.376% to 0.4% on which Dr Epstein relied, the rate set in the 2MMM case related to an FM radio station which has a far higher music use percentage than AM stations. The PPCA-CRA Broadcast Agreement applies to all stations. The rates thus reflect the decision in the 2MMM case which itself reflects the 1% cap.

196    Second, as to the fact that PPCA has not terminated the PPCA-CRA Broadcast Agreement, Dr Epstein did not appreciate that by s 109 of the Copyright Act there is a compulsory licence of sound recordings for use in broadcasts. Accordingly, Dr Epstein’s evidence that PPCA would be free to stop radio stations from using sound recordings in broadcasts and could have been expected to do so if it believed that the rates in the PPCA-CRA Broadcast Agreement were unreasonably low after the agreement expired in 2003 is misconceived.

197    Third, Dr Epstein’s reliance on the fact that no radio station is paying close to the 1% cap, which suggests that PPCA could have sought to renegotiate the agreement after its expiry in 2003 (there being, as he put it, plenty of “upside” available to PPCA worth a lot of money), does not indicate that the 0.4% is a market rate. The collection and distribution of licence fees by CRA members under the agreement is conducted by CRA. Until this proceeding PPCA did not know the internal division of liability as between CRA’s members and thus reasonably assumed that as some radio stations would be paying less than 0.4% due to their low music use percentage, others must be paying more and perhaps close to the 1% cap. The potential “upside” was thus not known to PPCA until recently. Further, and in any event, as this matter has demonstrated, seeking a determination by this Tribunal (unfortunately, as matters presently stand) involves an extraordinary amount of time, effort and cost. In circumstances where the best it could ever do was 1%, it made little sense for PPCA to try to renegotiate the 0.4% rate and even less sense for it to return to the Tribunal about that issue for so long as the 1% cap remains in place.

198    It follows from this that at the time they were negotiated the rates in the PPCA-CRA Broadcast Agreement were anything but freely determined market rates. They were rates determined within the constraint imposed by the 1% cap and in light of the broadcast right extending only to protected sound recordings. The fact that the rate of 0.4% has continued to roll over since expiry of the agreement in 2003 is also a result of the constraint imposed by the 1% cap. The 1% cap means that there is no incentive for PPCA to seek to renegotiate the rates irrespective of the extraordinary changes in circumstances wrought by digital technology and thus use of sound recordings since 2000. The 0.4% rate is best understood as an historical artefact. The fact that Dr Epstein and Mr Samuel were not informed about all of the relevant background circumstances to the 0.4% rate in the PPCA-CRA Broadcast Agreement undermines the value of their evidence about this issue.

199    Fourth, and reflecting a point Mr Massarsky made, there is another fact that is obvious about the 0.4% rate. It was fixed in 2000. The fact that it has rolled over since 2003, when everything in the music industry has been revolutionised over the same period by reason of digital technology, is an indicator that the 0.4% rate cannot be taken at face value. PPCA pointed out that in Australasian Performing Right Association Ltd v Federation of Australian Radio Broadcasters Ltd [1999] ACopyT 4; (1999) 46 IPR 20 the Tribunal said (at [13]):

A reference to a going rate is not a reference to a rate for a licence on quite different terms, or granted in significantly different circumstances, or relating to a changed product. The same logic demands that reasonable contemporaneity is also implicit in the adoption of such a rate. As time passes, a rate once adopted may require reappraisal, if not revision. It cannot continue indefinitely to be taken as the measure of what freely bargaining parties would necessarily now agree.

200    In the ordinary course, but for the 1% cap, one would have expected the 0.4% rate to have been revised over the period of 12 years since 2003 and the expiry of the PPCA-CRA Broadcast Agreement. The fact that PPCA has not bothered to do so exposes that market forces are not in play in respect of the broadcast licences between PPCA and radio stations. It is worth identifying the evidence Mr Massarsky gave in this regard, the cogency of which is manifest in respect of this issue:

And the second thing I wanted to mention, and it has to do with Dr Epstein’s mention that, well, you know, this .4 per cent – it has kind of been out there for a long time, and nobody seems to have bothered with it, maybe that means that it was an appropriately market-based rate. You know, the truth is that our entire music industry has changed in the last three or four years. As evidence shows by both economists representing PPCA, we see a tremendous transformation of an industry that used to rely on sales, from the sale of a record, to access models, the permission required to access intellectual property. I wasn’t here in Australia when those rates were settled on 10 years ago, but I bet you that the record companies in Australia, just like the United States, were very focused on the sale of recorded media.

Today everything counts. Everything is now under inspection. Those older deals are no longer valid deals or valid benchmarks. For the record industry now, that has become reliant upon subscription models and digital licensing models and what we still call access models for the – not custody of protected sound recordings, but rather the renting of sound recordings. And it just means that we have to look at ..... you have to proffer a decision on this case, you know, looking at it in the eyes of 2015, and you’re looking at the market realities of how the radio industry and the record industry are now interfacing with one another. It almost doesn’t make any rational sense to rely on a deal that was decades old and assume that because there wasn’t much objection to it, that it has a place in today’s music industry economy.

201    Fifth, and for what it is worth, Dr Eisenach’s review of rates for equivalent rights around the world resulted in the conclusion that the 0.4% rate was by far the lowest rate paid. While we do not know the constraints applying in all of those other jurisdictions, the fact that no-one could identify any country which recognises the broadcast right having set a lower rate than that in Australia is of some relevance.

202    Sixth, the fact that the rates set in the APRA-CRA Agreement are much higher than the 0.4% rate is also relevant. We deal with this agreement separately but, for present purposes, it is sufficient to note that in respect of the copyright for the musical work, where no 1% cap applies, APRA and CRA have negotiated the payment of far higher percentage of revenue rates than 0.4%. Dr Eisenach rightly considered this also to be a relevant indicator of the unreliability of the 0.4% rate, saying:

… the fact that APRA’s compensation for rights that closely correspond to the rights being licensed from PPCA is so much higher than what PPCA receives is itself an indication that the Broadcast Rate is not a market rate.

203    We note in this regard that in Mr Massarsky’s view because the risks and effort associated with the creation of a musical work are far less than the risks and effort to which a record company is subject in developing a sound recording, the sound recording right is in fact worth more than the musical works right.

204    CRA’s submissions about the 1% cap and 0.4% rate were unpersuasive. Insofar as those submissions have not been addressed in the matters set out above, we make the following points.

205    PPCA has not “conducted itself commercially in a manner inconsistent with the notion that it believes at 0.4% it is constrained by the cap”. The reasons CRA gave in support of this submission fail to recognise the fundamental fact that the best PPCA could do is 1% of revenue and, based on the history of the 2MMM case, it was reasonable for PPCA to believe that it would likely achieve less than 1% due to the broadcast right being confined to protected sound recordings. CRA also overlooks the enormous time, effort and cost that would be involved in referring the broadcast rate to the Tribunal for potentially little reward given the 1% cap. The fact that PPCA instead focused its resources on getting the 1% cap removed from the legislation is commercial conduct consistent with PPCA’s view that the 1% cap constraint operates to suppress the rate that it believes should be paid for sound recordings.

206    PPCA’s conduct in respect of Foxtel, where PPCA terminated its agreement because the rates paid were too low, says nothing about PPCA’s view of the 1% cap. Foxtel does not have the benefit of s 109 of the Copyright Act whereas radio stations do have the benefit of that provision. PPCA, accordingly, cannot prevent radio stations from using sound recordings in their broadcasts. This creates a different commercial dynamic when it comes to any possible termination of the PPCA-CRA Broadcast Agreement.

207    While it is true that, in theory, PPCA could seek from the Tribunal the imposition of a rate of 1% of revenues irrespective of the distinction between protected and unprotected sound recordings, PPCA’s view that this is not the way the Tribunal approached the issue in the 2MMM case is reasonable. If the distinction had been irrelevant in that case, there would have been no need for evidence to be adduced about the proportions of protected and unprotected music played on 2MMM. PPCA is right to infer that the evidence was adduced because 1% was seen as the upper limit and the rate to be selected was seen as properly reflecting the percentage use of protected recordings only. PPCA had reason to believe that CRA would advocate for the same approach in any further reference to the Tribunal, with the result that the “upside” available would not be worth the time, effort and cost involved to achieve any increase on the 0.4%. This also answers the submission which CRA made that:

Despite there being a compulsory licence, Dr Epstein is correct that PPCA could choose to revoke the rights from CRA's members at any time. The result would be the broadcasters commencing Tribunal proceedings to obtain orders for a new scheme. However, in order to get to that point, there would have to be an unsuccessful negotiation first.

208    Why, it might be asked, would PPCA bother to involve itself in such a process when it would involve such time, effort and cost for most likely little possible reward.

209    CRA’s repeated references to the 1% cap applying only to the Tribunal and not to negotiated rates are without merit. No radio station would agree to pay more than 1% in circumstances where the only available means to resolve a deadlock, the Tribunal, cannot impose more than 1%. This also answers CRA’s repeated submission that:

It is notable that PPCA has not even attempted to renegotiate the licence in 13 years.

210    This also undermines the evidence of Dr Epstein on which CRA relied that:

I don’t see anything preventing PPCA from attempting to renegotiate the 2000 agreement. It only was scheduled to run until 2003, so the expectation, I would imagine, in 2003, was that it would be renegotiated. It was not. But there’s nothing to stop the rate from being renegotiated and going to some other value, and all I can say is that the effort was not even made.

211    Effort is only made if the effort offers a potentially meaningful reward when compared to the other options available. PPCA’s decision to allow the 0.4% rate to roll over and to focus instead on the removal of the 1% cap makes sense. It represents the actions of a rational economic actor faced with a significant market distortion about which it has no practical options available other than to attempt to get the distortion removed.

212    For these reasons the approaches of Dr Epstein and Mr Samuel which take the 0.4% rate as their starting point for the value of the simulcast right are misconceived. These approaches cannot provide a reliable foundation for assessing the value of the simulcast right expressed as a percentage of revenue.

5.4    The APRA-CRA Agreement

5.4.1    General matters

213    The APRA-CRA Agreement commenced on 1 January 2010. It grants to each radio station which takes up the licence the right to:

Communicate the APRA Works in Australia by means of Broadcasts by the Station (which, for the avoidance of doubt, includes simulcasts from the Website).

214    As PPCA explained in its written submissions:

Radio stations are currently licensed to broadcast and simulcast musical works pursuant to agreements referred to in the experts’ joint report as the “APRA-CRA Agreement.” In fact, each radio station has a separate agreement with APRA, which also makes the agreement as agent for AMCOS. … Of course, it is clear that CRA negotiates the agreements with APRA; they are all in common form. Indeed, it would be ludicrous to suggest that an individual radio station/APRA|AMCOS agreement represented a voluntary bargain between those parties: it is a bargain struck between CRA and APRA|AMCOS.

215    The percentage of revenue rates payable under the APRA-CRA Agreement are as follows:

216    As PPCA pointed out:

The amount paid varies according to the station’s MUP, but a station with an average MUP (61%), pays 2.42% of revenue to APRA, about six times the 0.4% of revenue payable pursuant to the PPCA-CRA Broadcast Agreement

217    While Mr Samuel made use of the APRA-CRA Agreement in his alternative calculations, only Dr Eisenach considered that agreement to be relevant in that it showed what radio stations in Australia were willing to pay for both the broadcasting and simulcasting rights for musical works. Insofar as he dealt with this agreement, Mr Massarsky’s opinion was that the sound recording right was worth more than the musical works right because of the risk and reward structure being far more fraught for sound recordings. He pointed to a 12:1 ratio on the rates Pandora paid for sound recordings compared to musical works as supporting this opinion. He also pointed to a deal between Universal and Rdio which indicated a far higher percentage of revenue rate for sound recordings compared to that paid for musical works. PPCA noted that APRA had implicitly acknowledged a value ratio of 1:1 between the two rights in the Nightclubs case.

218    Because Dr Epstein and Mr Samuel based their opinions about comparable bargains on the PPCA-CRA Broadcast Agreement they did not treat the APRA-CRA Agreement rates as an indicator of comparability.

219    Apart from the 1% cap, we are unable to discern in the evidence any rational basis why the value of the sound recording right (broadcast and simulcast together) would be so much less than the value of the music works right (broadcast and simulcast together) in Australia.

220    For example, on the approach of Dr Epstein and Mr Samuel the combined value of the broadcast and simulcast right would be between 0.434% and 0.465% for sound recordings. For musical works, however, the combined value of the broadcast and simulcast right is between 0.054% to 3.76% with the lower end of this range for music use percentages of 0% - 9.99%. If an average is taken, the musical works rights in Australia would be worth about 4.5 times as much as the sound recording right (1.98% of gross revenue for the musical works right compared to 0.45% of gross revenue for the sound recording right) on the approaches of Dr Epstein and Mr Samuel.

221    CRA submitted that:

330.    It is well recognised that APRA’s repertoire is broader than PPCA’s.

331.    The inference from the PPCA record label witnesses that due to the internet and phenomena such as YouTube, people are now producing their own music without the involvement of the record labels, is that PPCA’s repertoire has decreased and there is less valuable.

332.    Further, stations use only a small proportion of PPCA’s repertoire that they licence.

222    We do not consider the inference in para 331 of CRA’s submissions to be reasonably open. As to para 332, this is a choice made by radio stations and we do not see it as material to any difference in value between the musical work right and the sound recording right. In respect of para 330, the evidence Ms McGregor gave was limited. She said only that PPCA’s repertoire was limited to protected sound recordings, but this is not the case for simulcasts (in contrast to the broadcast right). She said also that the musical works right extends to public performances of the musical work, but we cannot see this fact as material. PPCA submitted, and it was not disputed that, “PPCA had about 1581 member licensors covering a vast recorded repertoire. They include the majors Sony Music, Universal and Warner and around 21,000 licensor labels”.

223    The total value which Dr Epstein and Mr Samuel reached for the sound recording right is substantially below the total value of the musical works right. It could be expected that the evidence would disclose some rational reason for the difference. The evidence discloses nothing but the 1% cap. It does not disclose any other reason why the sound recording right would be of any less value than the musical works right. Indeed, based on Mr Massarsky’s evidence, it might be expected that the sound recording right would be worth more than the musical works right. Other evidence dealt with in more detail below, for example, equivalent rates in Canada, New Zealand and the UK, support the view that there is no material difference between the value of the two rights.

224    These considerations support our view that it would be unreasonable to use the 0.4% rate from the PPCA-CRA Broadcast Agreement as the foundation for valuing the simulcast right. They also indicate that the parties, because of their entrenched positions, have failed to give sufficient weight to the APRA-CRA Agreement as a relevant bargain. The APRA-CRA Agreement was entered into on behalf of the same radio stations which are one half of the notional bargain in the present case. APRA is a collecting society, as is PPCA. The APRA-CRA Agreement is still in force. There is no suggestion in the evidence that either party to the APRA-CRA Agreement considers it to be other than the result of a commercial negotiation. Unlike the PPCA-CRA Broadcast Agreement, there is no 1% cap in play. CRA and APRA, therefore, may be inferred to have reached a bargain unaffected by legislative constraint. The bargain, moreover, concerns the musical works right which is a complement to the sound recording right. Radio stations need both rights to play music in their programs. Further, the issue of protected and unprotected sound recordings is irrelevant to the APRA-CRA Agreement, just as it is to the simulcast right for sound recordings.

225    All of these factors carry substantial weight in favour of making use of the APRA-CRA Agreement as a foundation for valuing the sound recording simulcast right. The value of the complementary right for musical works is embedded in the rates agreed by APRA and CRA.

226    We return to this issue after considering Mr Samuel’s alternative calculation which makes a different use of the APRA-CRA Agreement.

5.4.2    Mr Samuel’s alternative calculation

227    In response to the Tribunal’s request that Dr Eisenach, Dr Epstein, Mr Massarsky and Mr Samuel propose alternative methodologies assuming their principal methodology was not accepted by the Tribunal, Mr Samuel put forward an alternative calculation not using the PPCA-CRA Broadcast Agreement rate of 0.4% as the starting point.

228    CRA described this alternative as follows:

132.    The calculation is essentially as follows:

(a)    using the licence fees paid by CRA’s members to APRA|AMCOS under the Current APRA Agreement in FY11 and FY12, and the uplift in rates between the Previous APRA Agreement and the Current APRA Agreement (which was on average 7.6%), calculate the notional licence fee which would have applied under the Previous APRA Agreement in FY11 and FY12 by subtracting the uplift amount from the actual licence fees paid;

(b)    deduct from the licence fees paid by CRA’s members to APRA|AMCOS under the Current APRA Agreement:

(i)    the notional licence fee calculated for each of the relevant years; and

(ii)    the licence fees that were paid by CRA's members to AMCOS under the Previous AMCOS Agreements;

(c)    apply a percentage figure that represents the proportion of the uplift amount referable to the simulcast right, given that the Current APRA Agreement includes 5 new communication rights which were not included in the Previous APRA Agreement and the Previous AMCOS Agreements, being (in addition to the broadcast right and the other AMCOS reproduction and communication rights which were included in the previous agreements):

(i)    a simulcast right;

(ii)    a streaming right (including on-demand streaming, except as part of an interactive on demand music service);

(iii)    a webcasting right;

(iv)    a podcasting right; and

(v)    an additional AMCOS communication right.

Mr Samuel’s view is that the proportion of the bundled rights to be attributed to the simulcast right is "pretty nominal". Allowing 20% in all the circumstances is a reasonable approximation, being 1/5th of the new bundled rights. The amount paid by CRA members under the APRA|AMCOS Interim Simulcast Licence also suggests that 20 % is a reasonable estimate. On one view, it is overly conservative, given the independent economic activities able to be engaged in using some of the other rights, for example, on-demand streaming and podcasting; and

(d)    determines the royalty as a percentage of CRA members' gross revenue in FY11 and FY12 by dividing the proportional notional licence fee calculated under paragraph (c) above by the CRA members' gross revenue for those years.

133.    The Samuel Further Calculation results in a royalty rate of 0.029% of gross revenue, assuming 20 % of the royalty is attributable to simulcasting which is Mr Samuel’s reasoned and "conservative" (in favour of PPCA) opinion but may be adjusted.

229    While Mr Samuel attempted to meet the Tribunal’s request we consider the attempt unreliable because it assumes, amongst other things, that the simulcast right was not covered by the previous APRA-CRA Agreement. We found CRA’s submissions that the parties to the previous APRA-CRA Agreement must have believed that the simulcast right was not covered difficult to follow. CRA, one party to the agreement, believed that the simulcast right was part of the broadcast right. It seems that the only party who held a contrary view was PPCA. It was not until the decision of the Full Court in 2013 that PPCA was proved correct. Until that moment the available material shows that PPCA was effectively on its own – everyone but PPCA was operating on the assumption that the broadcast right included the simulcast right. It follows that Mr Samuel’s assumption that the previous APRA-CRA Agreement did not licence simulcasting is without foundation. The parties to that agreement must be taken to have assumed that the agreement also authorised simulcasting. The rate in the previous APRA-CRA Agreement was thus taken by the parties to apply to both rights.

230    History discloses CRA’s firmly held view that the broadcast right included the simulcast right. This is sufficient of itself to dismiss the suggestion that CRA and APRA, before 2010, negotiated an agreement which they understood thereafter not to include simulcasting.

231    It is also relevant that when they negotiated the current APRA-CRA Agreement the parties defined the relevant right as the right to “[c]ommunicate the APRA Works in Australia by means of Broadcasts by the Station (which, for the avoidance of doubt, includes simulcasts from the Website).” This wording conveys a common belief on the part of the parties that, PPCA having raised the issue by this time, CRA and APRA at least were satisfied that the broadcast right granted by APRA in previous agreements automatically included the simulcast right. This description, it must be inferred, was included because PPCA was asserting to the contrary. The description is a strong indicator that APRA and CRA believed the broadcast right included the simulcast right.

232    CRA placed substantial weight on the fact that in 2002 CRA and APRA negotiated rates for simulcasting. Two examples of radio stations having entered into simulcast licences at these rates are in evidence. The rates in both licences involve a flat quarterly fee of less than $315. In other words, the rates to be paid are insubstantial. These licences should be understood in the light of the position that CRA adopted at all times before the High Court refused special leave from the Full Court’s orders that the broadcast right included the simulcast right. In that light, these simulcasting licenses, represent an abundance of caution between CRA and APRA negotiated on the basis of an insubstantial payment.

233    This is also consistent with CRA’s stated position after the refusal of the special leave application. In its submissions to government on the issue CRA said:

(a) The issue before the Committee is a simple one. The Broadcasters ask that the Minister immediately makes a determination in the terms set out in Paragraph 12 of the Initial Submission. This determination will reinstate the status quo as understood and applied by broadcasters over a long period, without dispute by PPCA or any other organisation.

(b) The 'status quo' erroneously mentioned by the Attorney General’s Department is actually the position adopted by all participants in the industry up until 2009 and by all participants in the industry (other than PPCA) and Justice Foster of the Federal Court until the decision of the Full Court of the Federal Court on 13 February 2013 in Phonographic Performance Company of Australia v Commercial Radio Australia Limited [2013] FCAFC 11 (see paragraph 8 of the joint submissions of CRA, CBAA, the ABC and SBS dated 10 May 2013). It is not correct to say that the 'status quo' is the decision of the Full Court made less than 6 months ago.

(c) The New Interpretation found that the simultaneous online communication of a broadcast was not itself a broadcast. This is contrary to longstanding and accepted practice on which most broadcasters have relied for many years.

234    In these circumstances the simulcast licences between APRA and two radio stations cannot be taken as evidence that CRA and APRA understood that the agreements before 2010 did not include the simulcast right. If anything, these licences appear to be an anomaly given CRA’s otherwise consistently expressed position that it and everyone else but PPCA accepted that the broadcast right included the simulcast. This might explain why only two licences are in evidence, there being no suggestion that all or any other radio stations entered into such a licence. It follows that we do not accept CRA’s submission that implicit in this conclusion (being the conclusion that Dr Eisenach reached when the simulcast licences were disclosed to him):

is the absurd suggestion that serious commercial parties (CRA and APRA|AMCOS) had entered into an additional agreement to licence a right which has already licensed under a separate agreement and for which CRA's members were already paying a fee.

235    We do not consider Dr Eisenach’s conclusion absurd; we agree with it. The entry into the simulcast licences at insubstantial payment rates, given CRA’s otherwise consistently expressed views about the broadcast right including the simulcast right, is best explained as an exercise of abundant caution by the two radio stations involved.

236    Otherwise, in respect of CRA’s submissions on this issue:

(1)    It is not to the point that any previous APRA-CRA Agreement pre-dates the commencement of simulcasting in 2001. The point is that on the available material CRA and APRA both considered that the agreements in place authorised the simulcasting activity once that activity commenced.

(2)    It is not the case that “on no interpretation would parties enter into a separate agreement if they believed the rights were already covered in a pre-existing agreement”. Parties frequently enter into agreements and subject themselves to additional provisions as a matter of abundant caution. Provided the additional provision is not too onerous there is no disincentive to parties doing so. In the present case, APRA and CRA must be taken to have shared the view in 2002 that while they believed the existing agreements covered the simulcast, another agreement, providing for payment of a small amount and imposing no additional reporting requirements, would not be inappropriate.

237    It follows that it is not possible to place any reliance on Mr Samuel’s alternative calculation. All that comparing the earlier APRA-CRA agreements with the current APRA-CRA Agreement discloses is that the rates increased over time. The average increase, according to Mr Samuel, was 7.6%. The fact that these rates increased by 7.6% over about a 10 year period whereas the rate under the PPCA-CRA Broadcast Agreement has remained at 0.4% since 2003 is another factor supporting our conclusion that the 0.4% is not a valid starting point for the calculation of the value of the simulcast right.

238    Otherwise we accept PPCA’s submission that Mr Samuel’s alternative calculation has problems, even if it be assumed that (contrary to our opinion) the fundamental assumption upon which it relies has validity. Adopting PPCA’s submissions in this regard:

(1)    Mr Samuel did not consider that the relative value of the broadcast right most likely decreased over time proportional to the increased audience percentage listening via the simulcast. Mr Samuel assumed the broadcast right had a constant value.

(2)    Mr Samuel’s allocation of value to the simulcast right compared to the other new rights in the APRA-CRA Agreement was (as he acknowledged) arbitrary.

5.5    Canada, New Zealand and the UK

239    Schedule D to CRA’s submissions summarised the material available in respect of equivalent payment schemes in Canada, New Zealand and the UK.

5.5.1    Canada

240    It is apparent from that summary that Canada has adopted a view that the sound recording and the musical works rights have the same value. As CRA’s summary put it:

It has been settled law in Canada that there is no reason to believe that the sound recording right has a greater value than the value of musical works. This is contrary to the proposition that appears to have been put by some of PPCA's witnesses that the sound recording right is valued more highly than the musical works right in the US. The Canadian Copyright Board has long held that SOCAN [the Society of Authors Composers and Publishers that represents authors, composers and publishers for the public performance of their music in Canada] and Re:Sound's [a not-for-profit organisation that represents performers and sound recording makers for the public performance of published sound recordings in Canada] tariffs are set at a 1:1 ratio, subject to repertoire adjustments (usually determined on the basis of music use studies).

241    The Canadian Copyright Board rejected a per stream payment system for simulcasting on the basis that simulcasting was ancillary to broadcasting for which radio stations paid a percentage of revenue to SOCAN and Re:Sound. The Board determined initially that the rate for simulcasting should be 2.1% of internet related revenue (SOCAN Tariff 22.B). Subsequently, in 2014, in Re:Sound Tariff 8 the Board rejected an application by Re:Sound for a per steam rate for simulcasting on the same basis, that the value of the simulcast should be determined in conjunction with that of the broadcast, the relative values of each being proportional to the relative audience. It deferred setting rates for simulcasting until its next consideration of the broadcast rates (an option not available to this Tribunal given the 1% cap). In respect of webcasting, for which Re:Sound had also sought the setting of a tariff, the Board set a general rate (subject to some exceptions) far lower than that sought by Re:Sound (which had based its rates on those applicable in the US), being $0.000102 per stream (0.0102 cents) included in a webcast, with a minimum fee of $100 per year. In so doing the Board concluded that there were significant differences between the markets in the US and Canada including that:

    The regime in the US was specifically directed at preventing the diminution in record sales because of digital piracy which was not the object of the legislation in Canada.

    “… American rate setting has been affected by a series of unique political and economic dynamics, steeped in controversy that has led to intervention on the part of Congress.”

    “In addition to American rates, the parties supplied us with a variety of other foreign rates for webcasting sound recordings … these rates vary greatly from country to country, with the American rates being the highest. There is no reason to believe that American rates are any better a proxy for our purposes than other foreign rates”.

5.5.2    New Zealand

242    In New Zealand in 2010 the New Zealand Copyright Tribunal determined rates for broadcasting and simulcasting (Phonographic Performances (NZ) Limited (PPNZ) v Radioworks Limited and The Radio Network of New Zealand Ltd (Radioworks) [2010] NZCOP 2 (PPNZ v Radioworks)). Under this determination:

    The rate payable by commercial radio broadcasters in respect of the broadcast of PPNZ sound recordings for other than low music use stations is 2.6% of Broadcast Gross Income in the first year of the licensing scheme and 3% of Broadcast Gross Income in the remaining years of the licensing scheme.

    For low music use stations, the rate payable in respect of broadcasts is 0.5% of Broadcast Gross Income in the first year of the licensing scheme and 0.75% of Broadcast Gross Income in subsequent years.

    The rate payable in respect of simulcasting should be calculated as a percentage of revenue, the rate being the same percentage of revenue as determined for the communication of sound recordings via the broadcast, but adjusted downward by multiplying the ratio of “audio page impressions” (being visits to the pages of a station's website on which the simulcast could be accessed) to the total page impressions across a station’s entire website.

243    In 2013 PPNZ entered into an agreement with The Radio Network. The primary difference between this agreement and the 2010 determination is that simulcasting is not paid for by reference to the concept of “audio page impressions”. Simulcasting is instead incorporated within the other rates payable under the 2010 determination (3% of gross revenue for other than a low music use station).

244    In New Zealand there are similar rates for musical works. For stations with over 50% music use, the rate is 2.9% of gross revenue. For radio stations with less than 50% music use, the rate is 1% of revenue.

5.5.3    The UK

245    In the UK simulcasting is subject to the same percentage of revenue payment system as broadcasting, whereas webcasting is subject to a per stream payment system. The initial broadcast rates determined were 5%, 3% or 2% of “net broadcast revenue” (or NBR), depending on the level of revenue generated by the radio station, NBR meaning 85% of revenue, a deduction of 15% being made for advertising-related expenditure. In 2001, by agreement between the relevant collectives, the definition of NBR was amended to include advertising from simulcasting, but the percentages payable of between 2% and 5% depending on revenue remained unaltered. This remained the case in 2006 when the relevant agreement was again re-negotiated. The position with respect to the use of musical works is not dissimilar. The applicable rates for use of musical works are 5.25%, 4% and 3% of NBR. Simulcasting was included by amending the definition of NBR without alteration of the percentage rates.

246    In respect of webcasting in the UK, a per stream payment system applies. The UK Copyright Tribunal in 2007 described the services of simulcasting and webcasting as “quite different” (British Phonographic Industry Limited and Ors v Mechanical-Copyright Protection Society Limited and Ors, CT84-90/05, 19 July 2007, 16 at [148]). As CRA summarised:

Having determined that it would treat simulcasting in the “offline category” with broadcasting “even though available via the Internet”, the UK Copyright Tribunal proceeded to identify a number of differences between the offline services (being commercial radio broadcasting and simulcasting) and online services (being webcasting). This included, for example, what the Tribunal described as differences in output (that webcasting is a pure music service of largely uninterrupted music streams and the large number of truly genre-specific offerings - whereas commercial radio includes a range of music, presenter speech, news, information and entertainment, all of which require significant additional resources); the regulatory requirements of commercial radio broadcasters (which are not required for webcasting services); that webcasters do not add value to the output in the way a commercial radio broadcaster does, through the voice of a popular disc jockey, the programme content, the format etc; the high level of control and interactivity provided to listeners of webcasters which are not available to listeners of broadcast or simulcast radio; and the high costs of setting up and operating a broadcast radio station compared to the costs of setting up and operating automated online services.

5.5.4    Observations

247    The following propositions are supported by the available material in respect of these three jurisdictions:

(1)    In Canada, New Zealand and the UK the sound recording right is treated as having a value about equivalent to the musical works right.

(2)    The rates in each country for the musical works right are reasonably in line with the rates in Australia for that right. In Australia the range of rates for the musical works right is between 0.054% and 3.76% of gross revenues depending on music use percentages. In Canada it is between 1.5% and 3.2% of the first CAD$1.25 million of gross income and 4.4% of the rest of the gross income of the radio station. In New Zealand it is between 1% and 2.9% of gross revenues depending on music use percentages. In the UK it is between 1% and 5.25% of NBR.

(3)    The rates in each country for the use of sound recordings in radio broadcasts are well above the 0.4% rate in Australia. In Canada the rates are between 0.75% and 2.1% depending on music use. In New Zealand the rates are between 0.75% and 3% depending on music use. In the UK the rates are between 1% and 5% of NBR depending on music use.

(4)    Canada, New Zealand and the UK have not adopted a per stream rate for the simulcast right. In each of these countries (like Australia but unlike the US) there was, before the simulcast right became possible by reason of technology, a royalty payable for the use of sound recordings and musical works in radio broadcasts which was and would continue to be calculated on a percentage of revenue basis. When simulcasting occurred, the relevant bodies in each country decided that simulcasting, unlike webcasting, should also be compensated for on a percentage of revenue basis in conjunction with the broadcast right. In each country the broadcasting rate appears to have been determined on the basis of notions of fairness and reasonableness. There is no indication that the rates set for the broadcast right in these other countries have been subjected to legislative direction. This has enabled the simulcast right to be valued by reference to the value fixed for the broadcast right.

248    We consider that these propositions support our conclusion that the 0.4% rate is not a market rate but represents a significant distortion resulting from the 1% cap, the immediate practical consequence of which is that the simulcast right cannot be valued using the broadcast rate as a logical starting point (unlike the cases in Canada, New Zealand, and the UK). We also find in these countries support for our view that it is reasonable to treat the sound recording right as having a value similar to the value of the musical works right.

249    The result is that the current APRA-CRA Agreement provides a better guide to the value of the broadcast right and the simulcast right together than any attempt to make adjustments to the PPCA-CRA Broadcast Agreement. We deal with this issue next.

5.6    Using the APRA-CRA Agreement for percentage of revenue rate setting

250    PPCA contended that it was impossible to disaggregate the value of the simulcast right from that of the broadcast right in the APRA-CRA Agreement by reference to the notion of “an uplift” partly due to the simulcast right being newly granted in the 2010 agreement. We agree with this submission for the reasons given.

251    However, we are not presently minded to accept the submission that it is impossible to assign a reasonable proportion of the total value in the APRA-CRA Agreement to the simulcast right.

252    Dr Eisenach’s opinion was that the value of each of a bundle of rights cannot be disaggregated when a single price is attached to them. This is because if the licensee wants to purchase the broadcast right and the simulcast right, it considers the total price that it is willing to pay, and is indifferent about the apportionment of price between those rights. At the level of principle, this must be right. But our task is ultimately guided by what is reasonable in the circumstances. We do not consider it unreasonable to seek to identify the value of the simulcast right for musical works from the overall value, simulcast and broadcast, for musical works by any method that seems rational.

253    One possible method of doing so, which seems to us to be rational, is to use the average simulcast audience share. In circumstances where simulcasting accounts for 10.7% of the current radio audience on average we can see no reason not to begin the assessment of the value of the simulcast right on that basis. We accept that the average percentage of the audience listening via the simulcast is not the end of the assessment. This is because, as we have explained, the current value of the simulcast right is not limited to the audience share it attracts. It follows that reasonableness would also require an estimate of an increased percentage to account for those otherwise unaccounted for benefits.

254    Assume that the reasonable range to account for these additional benefits, in the current circumstances, is between 1% and 10%. If the sound recording right has the same value as the musical works right, the percentage of revenue payments for simulcasting, by reference to the APRA-CRA Agreement, would be as set out in the table below.

255    We include in this table the value of the simulcast right at 10% (the current audience share, rounded down), 15% and 20% of the total value to demonstrate the possible range of rates (and, incidentally, the extent of distortion inherent in the 0.4% rate for the broadcast right).

256    It will be apparent that at 10% the highest rate for simulcasting approaches the broadcast rate of 0.4% and at 15% and 20% many of the rates are above the 0.4% broadcast rate. This does not necessarily suggest that the rates in the above table for a percentage of revenue payment for simulcasting are too high. Rather, it confirms that the broadcasting rate of 0.4% is far below what would be a market rate for that activity by reason of the 1% cap.

257    For these reasons, the observations of the Tribunal in Reference by Australasian Performing Right Association Ltd under s 154 of the Copyright Act 1968 (1992) 25 IPR 257 at 267 – 268 that the result yielded must be considered against the background of the past cannot be used in the present case to indicate that the outcome is “extravagant”. That observation by the Tribunal assumed that the amounts previously determined represented a fair market rate. That assumption does not apply to the amounts radio stations pay under the present PPCA-CRA Broadcast Agreement.

258    While we recognise that this approach means that some radio stations (which do not have a high percentage of listeners via simulcast) will pay too much and others (which have a higher percentage of listeners via simulcast) will pay too little, this is an unavoidable consequence of a percentage of revenue payment system. In this regard, the methodologies of Dr Epstein and Mr Samuel both used a simulcast audience share of 10.7% which creates the same potential for inequity yet CRA proposed their schemes as reasonable.

259    We also recognise that an approach such as this will mean radio stations, relatively speaking, will be paying more for the simulcast right for sound recordings than for the broadcast right for sound recordings having regard to the audience share of broadcasting compared to simulcasting. As we have said, however, this is a result of the distorting effect of the 1% cap. The fact that copyright owners are not paid market value for the use of sound recordings in radio broadcasts is not a valid reason to reduce the rate payable for simulcasts of sound recordings. Moreover, nothing in the available material suggests that percentage of revenue rates within these ranges would be beyond the reach of any radio station or affect their viability. To the contrary, given that they are paying equivalent amounts for the musical works right, it must be inferred that radio stations are able to pay these amounts for these kinds of rights.

260    Two further questions arise for consideration.

261    The relevant circumstances are that, although CRA proposed that the scheme should be based on a percentage of revenue, the methods it used to determine the percentage of revenue payable were different from those discussed above. CRA’s principal methods involved use of the PPCA-CRA Broadcast Agreement as a starting point – an approach we consider unreasonable and destined to undervalue the simulcast right substantially. Mr Samuel’s alternative method used the APRA-CRA Agreement but not in a way we consider reliable. The material placed before us provides a cogent foundation for the conclusion that the sound recording right has a value at least equal to the musical works right, however, none of the schemes used this as a basis to determine the appropriate percentage of revenue rates payable. Moreover, unlike the prospect of a hybrid scheme giving radio stations the option of paying on a per stream or percentage of revenue basis, we did not raise this possibility with the parties during the hearing.

262    The first issue in these circumstances is that the parties have not had an opportunity to be heard about this approach to the calculation of reasonable percentage of revenue rates for the use of sound recordings in the simulcast. As such, it would be unfair to reach any final conclusion about the reasonableness of the approach or the results it yields without giving that opportunity to be heard.

263    The second issue is not as readily accommodated because it depends on actions of the parties, not of the Tribunal. We have raised above the question of the Tribunal’s power to vary a scheme. On one view, if (say) we adopted the percentage of revenue rates set out in the 15% column in the table above, we would be doing no more than increasing the percentage of revenue rates proposed by CRA. On another view, we would be substituting for CRA’s percentage of revenue scheme another percentage of revenue scheme altogether. Resolution of this question (and any question of power in respect of a hybrid scheme) would be resolved if, having reviewed these reasons, one or other of the parties was willing to refer to the Tribunal a scheme which gives radio stations the option of paying on a percentage of revenue or per stream basis and which, in respect of the percentage of revenue basis, adopts an approach as set out above or something similar to it.

264    One final observation is appropriate. We have ventured into these matters only because while we accept CRA’s submission that it would not be reasonable to force radio stations to pay for simulcasting on a per stream (and no other) basis, we also reject as unreasonable every approach CRA proposed for calculating a percentage of revenue rate. Left on this basis, we would have no option given the terms of s 154(4) of the Copyright Act other than to reject CRA’s schemes. This seems an unsatisfactory outcome which we are doing our best to avoid.

5.7    The NAB-SX Agreement

265    The rates in PPCA’s proposed scheme are based on the US NAB-SX Agreement which was entered into in 2009.

266    As noted, Dr Eisenach and Mr Massarsky both considered the NAB-SX Agreement a comparable bargain capable of application in Australia, in effect, without any adjustment of rates.

267    While we accept that the NAB-SX Agreement involves the correct right, the right to use a sound recording in a simulcast, we do not accept that it is directly applicable to Australian circumstances for a number of reasons.

268    First, and most obviously, radio stations do not pay for the broadcast right in the US. Accordingly, the evidence about the buoyant economic position of radio stations in the US irrespective of the rates imposed in the NAB-SX Agreement is not particularly helpful. Australian radio stations are in a different position. They pay for the broadcast right, albeit the 0.4% not representing a market rate for the reasons already given.

269    Second, whether or not the NAB-SX Agreement represented a voluntary agreement between the parties (another of the many issues in dispute between the parties which we do not find it necessary to resolve), the inescapable fact is that the rates payable are currently in dispute in the Webcaster IV proceedings. In those proceedings while Sound Exchange is seeking increased rates for 2016 – 2020 which appear to take the NAB-SX Agreement rates as a starting point, NAB seeks substantially lower rates of US$0.0005 for each year from 2016 to 2020.

270    It is true, as PPCA submitted, that the US Copyright Board has not yet determined the Webcaster IV proceedings. The result is thus unknown. But that said, it is nevertheless relevant that in those proceedings NAB (the CRA equivalent) has not accepted that the NAB-SX Agreement represents an appropriate starting point. While this might be forensic posturing (there is no way to know at this time), in these circumstances it cannot be said that the rates in the NAB-SX Agreement currently represent a consensus in the US as to a market rate for simulcasting on a per stream basis.

271    PPCA’s submission that, in all of the circumstances, the approach of NAB in the Webcaster IV proceedings is “ambitious” because it depends on the proposition that the NAB-SX Agreement was not one between a willing seller and willing buyer is one of those imponderable issues that, in truth, can never be resolved on the material in the present case. The relevant point, in our view, is that whatever else it might be, the NAB-SX Agreement does not currently reflect a consensus about a market rate for the use of sound recordings in simulcasts in the US.

272    These three matters are sufficient to support the conclusion that the NAB-SX Agreement does not provide a reliable foundation for the task of assessing the value of the simulcast right in Australia. If anything, using the rates in the NAB-SX Agreement is likely to overvalue that right because they have been negotiated free from any circumstantial constraint imposed by a need to pay for the broadcast right, and to pay the compliance and auditing costs associated with those payments, and in circumstances where the weight of the evidence indicates that the monetisation of simulcasting in the US is materially more advanced than in Australia. This conclusion is reinforced by the evidence of Mr Kott, who is in the business in the US of assisting radio stations to monetise their digital offerings. As PPCA’s submissions put it:

Mr Kott’s unchallenged evidence explained the advantages of internet radio - including that ads can be targeted, interactive, and can include a visual element, and that the audience can be reached via work computers, smartphones, tablets and social media. Mr Kott also described the different types of online radio ad units - in-stream audio ads with companion banners, pre-roll video or audio ads and display banners. Finally, Mr Kott described WideOrbit’s model for effective monetisation of internet radio, and two case studies of effective monetisation of streaming by small networks of regional stations in the US.

273    In other words, in the US radio stations are further advanced in the use of strategies to monetise their simulcasting activities. The evidence in respect of Australia indicates that radio stations in this country are not advanced at all in this regard. Leaving aside the argument of Dr Eisenach that they may not have had any incentive to make money for simulcasting because they have not had to pay for using sound recordings in simulcasts, the present value of a potential is necessarily less than the present value of something which, in part at least, is actually being realised. In the US, the potential value of simulcasting has been recognised and is being exploited. In Australia, the potential value of simulcasting is in the process of being recognised by some larger metropolitan radio stations but is largely unrealised. The stage of development of the exploitation of simulcasting in the US is thus more advanced than it is in Australia. It should be inferred that the rates payable under the NAB-SX Agreement reflect this fact.

274    The evidence of Dr Coker and others about the importance of websites to the radio industry does not lead to a different conclusion. Large metropolitan stations, in particular, have a clear focus on the importance of their website to maintaining and building audience share. There are, moreover, obvious potential synergies between the websites of radio stations and simulcasting. Insofar as CRA’s evidence was to the contrary, we are unpersuaded. What we do not accept, however, is the suggestion that Australian radio stations are currently effectively monetising simulcasting (over and above the fact that the simulcast audience share represents an equivalent proportion of advertising revenue because such revenue is related to overall audience share). Monetising websites (which is evident in respect of at least larger metropolitan radio stations in Australia) is one thing; monetising simulcasting is another. The potential for synergies between the two is obvious but the evidence indicates that, thus far, it has not yet been realised to any material extent in Australia. For example, there was no evidence that any material amount of targeted advertising via the radio simulcast was yet occurring in Australia (Mr Davies of ARN said it offered targeted advertising but it was not apparent that this was via the simulcast as opposed, for example, to website advertising). There was no evidence of what was described as programmatic advertising occurring (apparently, this involves the use of software to automate the sale of advertising, as opposed to the traditional process that requires advertising staff to negotiate, arrange and execute advertising). There was no evidence of differential advertising rates between the broadcast and the simulcast. There was evidence that most radio stations drew no distinction between the broadcast and the simulcast in their advertising strategies. There was evidence of little revenue from simulcasting pre-roll advertising.

275    It follows that PPCA’s submission that “a number of radio networks in Australia are now engaging in programmatic advertising, or in advertising that is customised for consumers based on available demographic information” was overstated. The evidence of Mr Davies from ARN, on which PPCA relied, was evidence focusing on future rather than existing potential. Mr Davies said that the information collected in order for a customer to gain access to the customised version of iHeart radio “could be useful” in offering services to advertisers if there was sufficient volume. There was no suggestion in the evidence that either condition had yet been satisfied. To similar effect was the evidence of Ms O’Connor of Nova. As PPCA itself put it, “Ms O’Connor of Nova also gave evidence that Nova was in the process of moving towards programmatic trading across all its digital platforms”. This highlights the magnitude of the difference between Australia and the US. In Australia, Nova, one of Australia’s largest and most sophisticated radio stations, is beginning the move towards programmatic trading. In the US, the progenitor of all of these forms of market development, the process is well advanced.

276    The paucity of evidence about the monetisation of simulcasting in Australia compared to the US was highlighted by PPCA’s reliance on a recent media release by SCA, another major player in the metropolitan commercial radio market. The media release, dated 29 April 2015, announced:

a partnership [by SCA] with Triton Digital, the leading technology provider for the audio industry, to bring the first programmatic audio ad exchange to Australia. The partnership will enable advertisers or agencies to buy digital audio advertising programmatically and to “leverage audience data to create more effective connections between brands and consumers”.

277    PPCA noted that:

SCA is one of Australia’s three large radio networks (along with Nova and ARN). The advertising spend figures produced by OMD [an advertising agency] show SCA had by far the highest digital advertising revenue from OMD bookings – several times higher than ARN or Nova. The table also shows that $44,416 was paid to SCA for programmatic advertising in the year to April 2015. The Tribunal can readily infer that among the radio networks, SCA is the market leader in effectively monetising its digital offerings.

278    The relevant point for the Tribunal is that Australia’s largest and most sophisticated radio station when it comes to monetising digital revenues is only now bringing the first programmatic audio ad exchange to Australia and earned the relatively insignificant amount of $44,416 for programmatic advertising in 2015. Compared to the situation Mr Kott described in the US, it is apparent that Australia is years behind in this sphere of activity.

279    It follows that we found the conclusions to the contrary reached by Dr Eisenach and Mr Massarsky about this issue unpersuasive. Neither, in our view, gave sufficient weight to the factors referred to above in reaching their conclusions. Dr Eisenach considered a range of financial and other metrics for consumer music usage which led to his conclusion that the US and Australian markets are substantially similar with respect to factors likely to affect the value of the simulcast right. The metrics which Dr Eisenach considered were radio advertising per capita, online advertising per capita, mobile advertising per capita, consumer use data, and other information about the simulcast market of which Dr Eisenach was aware. This other information, set out in Section IV of Dr Eisenach’s first report, concerned the relationships between the recorded music industry, the radio industry and the online music distribution industry.

280    We accept that all of the factors which Dr Eisenach took into account are relevant. Our concern is that he has not given any real weight to the other factors which we have identified above. In particular, we do not see how it can be the case that the simulcast right in Australia would have anything approaching the value in the US in circumstances where the potential value of that right has already been realised to a material extent in the US but remains largely unrealised in Australia.

281    The other matter which causes us concern is the difference in size between the US and Australian markets. Mr Samuel made this point noting that the US market is more than 17 times larger than the Australian market in terms of total radio industry revenue. Dr Eisenach did not dispute this but noted that Mr Samuel had not explained how this factor was relevant. It may be accepted that Mr Samuel did not do so. The relevant point, however, is that it might reasonably be inferred that a market which has a total revenue of 12.1 billion pounds in 2013 (the US) compared to a market which has a total revenue of 0.7 billion pounds (Australia) is very likely to be subject to different market forces. In particular, it might reasonably be inferred that in such a large market there are cost efficiencies and a critical mass in play which is likely to make it easier to monetise new technologies. As CRA submitted, there was force in Mr Samuel’s evidence that “the differences between Australia and the US include labour costs, property costs and economy of scale issues – for example, there is one network in the US which has 850 radio stations, which is three and a half times larger than the entire Australian market for commercial radio broadcasting”. Given the magnitude of the differences we consider it was more the responsibility of PPCA to explain why that factor was immaterial than for CRA to explain why it was material. On the evidence as it stands, we consider it likely the difference in market size is likely to be relevant to the value of the simulcast right and is likely to lead to an over-value being ascribed in Australia if the rates in the NAB-SX Agreement are used as the relevant starting point.

282    We also consider it likely that the rates in the NAB-SX Agreement are affected by the fact that the licence granted under that agreement covers more than simulcasting. In particular, it also covers webcasting by radio stations. Whether or not webcasting is being carried out under the NAB-SX Agreement or some other agreements as Dr Eisenach suggested, the licence PPCA is offering is more confined than that obtained under the NAB-SX Agreement.

283    We do not give any material weight to the submission of CRA that PPCA’s per stream rates are demonstrably unreasonable because they would result in a payment which Dr Epstein calculated to be about 13 times greater than the amount radio stations currently pay PPCA for the broadcast right. We do not give this weight because, as discussed, we consider the amount paid for the broadcast right to be artificially low by reason of the 1% cap.

284    For these reasons the US rates in the NAB-SX Agreement, insofar as they are representative of the current position in the US having regard to the uncertainties resulting from the Webcaster IV proceedings, are not directly applicable to Australia. They are likely to represent a substantial over-valuation of the simulcast right in Australia. The NAB-SX Agreement is useful, however, because it shows rates that we consider to be too high to be reasonable in an Australian context for the reasons we have given.

5.8    Webcasting agreements

285    PPCA also relied on recent webcasting agreements entered into by record companies. PPCA said that these recent webcasting agreements, which use rates similar to those in the NAB-SX Agreement, are a powerful indicator that the NAB-SX Agreement is a comparable bargain. PPCA also noted that in these agreements Australia appears as a “top tier” or “core” territory so that the rates payable in Australia are similar to those in the US which PPCA said “further supports the comparability of the NAB-SX Agreement and the US rates applicable under that agreement for the purpose of considering the Scheme”.

286    The webcasting agreements are between major record companies and music streaming services such as Samsung Milk, Guvera, Rdio and the like.

287    We consider that these agreements support our conclusion that, insofar as the use of sound recordings by radio in simulcasts is concerned, the rates in the NAB-SX Agreement are too high. As a result, it is not necessary to deal with CRA’s submissions that PPCA’s reliance on these agreements was unsatisfactory because they had been introduced at the “heel of the hunt”.

288    There was a great deal of evidence and many submissions on the part of PPCA in support of the conclusion that webcasting and simulcasting are comparable activities. We are unable to accept this submission.

289    First, webcasting is a new and relatively untested business over the longer-term. New entrants are entering and exiting the market with different business models at a pace unlike anything seen in the radio industry which is well-established and has enjoyed relative stability in terms of audience numbers and growth in profits over many years. As such, the relative bargaining power of a new webcaster compared to a collective of radio stations when dealing with a record company is not comparable. The collective of radio stations will have greater bargaining power than most webcasters. The two industries are not readily comparable for this reason alone.

290    Second, webcasters are in a much weaker bargaining position in any dealings with record companies because they have no service to offer without music. Radio stations use music to greater or lesser extents. Radio stations can adjust their offering, making music play a different role in their programming over time. Radio stations thus enjoy greater flexibility as to content than webcasters which have no flexibility as to their content offering when negotiating with record companies.

291    Third, the relative weakness of the bargaining position of webcasters is evident in the requirement in the agreements that they pay a per stream rate or a percentage of revenue rate whichever is the higher, where the percentage of revenue rates nominated exceed by many orders of magnitude the rates any expert suggested might be reasonable for radio stations to pay.

292    Fourth, the service offered by webcasters, even when made to seem more like radio (eg Samsung Milk), is really nothing like radio at all. No matter what degree of interactivity offered, it was apparent that all webcasters offer far more than the largely passive listening experience which characterises radio listening. Competitions offered during broadcasts or online by radio stations do not alter the fact that a listener has no real choice about what the radio station offers. All webcasters offer a degree of control and choice to the listener about the music to which they listen which is foreign to the listening experience of a radio station. The webcasting agreements might refer to certain streaming services as “non-interactive” but interactivity is a concept which operates across a broad spectrum. Even a non-interactive webcasting service, by the use of various mechanisms (be they software controlled or involving direct physical input by the listener), offer a form of individualised or tailored listening service which radio does not.

293    Accordingly, we do not accept PPCA’s submission that these services are “very close to radio”. We consider these services, even when justifying the description of “lean-back listening”, are offering something different from radio. Requiring songs to be announced and advertisements does not transform listening to a stream of music by a webcaster into an experience like listening to a radio station where the listener has to take the whole package in common with every other listener (subject only, in the case of simulcasting, to the potential for targeted advertising). The fact that listeners to a radio station can choose genres (such as easy listening, classical or the like) is not equivalent, for example, to the Samsung Milk experience where the capacity to select by genre is far more fine-grained, and all that is offered is music, rather than whatever entire entertainment package is offered by a radio station. As Dr Epstein said the webcaster services involve a “far more sophisticated” offering, at least in terms of user control and catering to user tastes, than listening to the radio.

294    We also accept Dr Epstein’s evidence that the rates in the webcaster agreements have to be understood as one element of a complex array of rates negotiated as a package between record companies and webcasters. The packages involved disclose higher rates for services with greater interactivity. As Dr Epstein explained the rates chosen at the non or semi-interactive ad funded end of the spectrum in these packages are designed to encourage the listener to buy their way out of having to listen to advertisements by paying for a subscription rather than using the free service. Dr Epstein rightly said that this dynamic, which must be relevant to the rates which have been set, does not apply to simulcasting by listeners to radio. We agree.

295    Nor do we think that the rates paid by Pandora are directly comparable. We accept PPCA’s submission that Pandora demonstrates why a simple valuation approach based on the net present value of present and future profits is not sufficient to value the simulcast right. Pandora, as Dr Eisenach and Mr Massarsky explained, is an enormously successful company which has not yet reported a profit. But the rates paid by Pandora do not show that a radio simulcast is equivalent or more valuable than the streaming service which Pandora offers despite Pandora’s rates being lower than those in the NAB-SX Agreement. Pandora offers a service unlike radio. As Dr Eisenach explained in the following exchange:

DR EISENACH: Pandora’s genius was to develop an algorithm which allowed it to identify if you listen to a song, based on what you listen to, to identify other songs that you would be likely to want…

…

Pandora allows you to put in an artist and it will automatically construct a radio station for you. And there are two flavours of Pandora. There’s a subscription-based flavour which is about five per cent of its subscribers and the non-subscription-based version which is about 95 per cent ad-supported. It is a non-interactive service. It’s often referred to as semi-interactive, but non-interactive in the sense that it’s limited to the number of skips advances. It doesn’t allow you to rewind. So many of the same kind of constraints that we saw for the new Webcaster agreements also apply to Pandora.

HER HONOUR: But it serves up the music it thinks you like?

DR EISENACH: It serves up the music it thinks you like. So I have several radio – I have several stations on my Pandora, so if I want to listen to Mozart it will play – Mozart-like stations. There are limits on – it won’t let you play the same artist – one of the limits that the record companies put on is that you can’t play the same artist over and over, so - - -

HER HONOUR: But you can select an artist – how many times - - -

DR EISENACH: So if I put in Bob Dylan, it will play Bob Dylan-like music.

HER HONOUR: Right. But can you keep putting in different artists and select - - -

DR EISENACH: .....

HER HONOUR: You can. And it just keeps modifying your station?

DR EISENACH: You can have as many stations as you want.

296    Other than this we note that the percentage of revenue rates Pandora is paying exceed anything that would occur to us as reasonable for radio stations to pay for the simulcast right. This indicates that the Pandora per stream rates are also too high for the simulcast right in Australia.

297    Fifth, we are not persuaded to reach any different conclusion because of what Dr Eisenach described as a striking feature of Universal’s webcasting agreements, that they contained prices for the same services in different territories – including Australia and the United States. Webcasters are a business model in rapid development in both countries. Dr Eisenach said that the agreements showed “directly comparable products, directly identical parties setting prices where the only difference is the two markets”. We consider however that the agreements are between different parties, with different relative bargaining positions in respect of a different product.

298    Sixth, we consider there to be substance in a point made by Mr Samuel that the business models of webcasters and radio stations are so different that their respective bargaining positions with PPCA (or record companies) must also be different. As Mr Samuel put it:

… none of these services is required to provide a broadcast.

Now, the consequence of that comes to one of my fundamental points, and something I’ve been pushing somewhat ad nauseam. This has to do with the financial consequences when you look at these agreements. If you don’t need to provide a broadcast, you don’t need to pay for transition towers, a radio licence, personalities at breakfast drive-time, traffic reports, news reports. So the cost structures of these entities is going to be fundamentally different to the cost structures of an Australian commercial radio broadcaster.

299    As Mr Samuel also explained, using CRA’s words, “the ‘greater of’ rates structures in the record label webcasting agreements means that the per stream rates will only apply if it results in a royalty that exceeds the percentage of revenue amounts - the minimum royalty is therefore the percentage of revenue amount”. In relation to the percentage of revenue figures, Mr Samuel said that it is “demonstrably uneconomic” for Australian commercial radio stations to pay a royalty which is anything remotely in that region. We agree.

300    Accordingly, while we agree with PPCA that much unnecessary effort was expended in this case about whether certain forms of webcasting are more, less or equivalently interactive to radio, the fundamental point is that webcasting is not sufficiently like radio in substance to make the webcasting agreements a useful starting point.

301    For these reasons we are unable to accept PPCA’s submission that we can “comfortably find that any minor differences between the interactivity of simulcasting and that of non-interactive webcasting is not so significant that the webcasting rates do not support PPCA’s proposed simulcasting rates”. In the circumstances we have described we are not persuaded that the bargains struck between the record companies and webcasters are a reliable guide to the notional bargain that would be struck between PPCA (or the record companies) and the radio stations which, unlike webcasters, have years of experience in the benefits of collective bargaining, are an established industry, offer a different entertainment package than webcasters of which music forms only one part, and have more flexibility in respect of music use than webcasters.

302    In common with the NAB-SX Agreement, all of the factors which we consider to be of weight demonstrate that the recent webcasting agreements disclose per stream rates that would be too high for the use of the simulcast right in Australia by radio stations. The agreements are therefore useful for this purpose.

5.9    The PPCA webcasting agreements

303    In 2013 and 2014, PPCA entered into a series of agreements with webcasters (some of which are already defunct) granting them the right to use sound recordings in their services, the required payment being on a per stream basis. The rates payable by these webcasters varied, but are substantially less than the rates in the NAB-SX Agreement (in general, somewhat less than half of those rates). PPCA expressly identified those agreements as “non-precedential”. It may readily be accepted that PPCA was, in effect, “testing the waters” with these agreements, the other parties to which were start-up companies entering a new market.

304    We accept, therefore, PPCA’s submission that these agreements are not comparable to the notional bargain that would be struck between PPCA and the radio stations. However, unlike PPCA, which wished to stress the lack of comparability of these agreements because they involve rates far less than the NAB-SX Agreement, we consider these agreements lack comparability for reasons similar to those set out above for the more recent webcasting agreements between the record companies and webcasters. Despite the fact that some of the webcasters had real commercial muscle behind them the enterprises in which they were involved were all experimental. They were all negotiating as start-up enterprises with no proven track record. The proposed enterprise was dependent on obtaining the required rights to use the sound recordings. The new enterprises had no flexibility in their use of music. Without music, they would not exist at all. These factors tend to suggest that the webcasters which entered into deals with PPCA in 2013 and 2014 were in a weak bargaining position in comparison to radio stations wishing to use sound recordings in their simulcasts.

305    While PPCA might have been willing at the time to offer these enterprises discounted rates compared to the NAB-SX Agreement to account for all of these factors, it is unrealistic of PPCA to assume that its bargaining position with radio stations would be the same. The NAB-SX Agreement provides an obvious appropriate starting point for negotiation with webcasters, as the deals the record companies have recently struck with webcasters disclose. But is does not provide an appropriate starting point for negotiations with radio stations for the use of sound recordings in simulcasts for the reasons set out above. As such, it is not realistic to characterise the PPCA webcasting agreements as embodying any form of discount insofar as those rates might be applied to radio stations. If radio stations had been negotiating with PPCA at that time, for the reasons we have given, they would have been in a position to negotiate lower rates than these start-up, experimental, enterprises. The PPCA webcasting rates are thus also useful in that they disclose per stream rates higher than what would be achieved in a notional bargain between PPCA (or the record companies) and radio stations.

306    Insofar as it is necessary to record this, we have difficulty with Dr Eisenach’s opinion that the agreements, taken together, show a trend of increasing rates for webcasting. We consider it impossible to discern any trend from the various agreements, each of which appears to reflect individual circumstances that happen to have pertained at the time. Unlike PPCA the recent agreements by the record companies were not experimental or a testing of the waters by the record companies. They represent a complex system of payments of which a per stream rate for ad funded so-called non-interactive services forms but one part. These facts alone might explain the difference in rates rather than an inference of a general trend in per stream rates for webcasting increasing.

5.10    Using the various agreements for per stream rate setting

307    CRA’s position, in common with the position of PPCA in the setting of a reasonable percentage of revenue rate, is that PPCA failed to demonstrate any rational foundation for a per stream rate. While we agree that none of the agreements which were put to us by PPCA provide an appropriate comparison for the assessment of a reasonable per stream rate for radio stations in Australia to pay for the use of sound recordings in simulcasting, we do not consider that the available material provides no rational basis for the assessment of a reasonable per stream rate.

308    Our starting point is that all of the webcasting agreements which we have identified above involve rates which exceed what would be reasonable for radio stations to pay for use of the sound recording right in simulcasting. This is because, as we have said, the businesses are different and have different cost structures, the products are different and radio stations are in a better bargaining position than webcasters in any (necessarily hypothetical) negotiation with PPCA (or the record companies).

309    Our difficulty, however, is that while we know from the available material rates that would be too high, we have not yet identified a rational way of determining rates that would be too low. We could estimate those rates based on a general conclusion that, in all of the circumstances we have identified, radio stations are likely to do substantially better than the webcasters PPCA did deals with, but this would still leave open a substantial range within which the selection of a particular rate would be guesswork. If necessary, we will not shy away from this task, guesswork or not, but will take into account any further submissions the parties may wish to make in this regard.

310    An alternative approach is provided by Dr Eisenach’s forecast of scheme fees payable to PPCA between 2013 and 2021 (Exhibit 94). The first observation we must make is that we are not certain we understand how the figures shown in this table have been derived. As such, and in common with our approach to the APRA-CRA Agreement, the following discussion is put forward on a provisional basis only, in circumstances where we are doing our best to avoid these proceedings being rendered futile. Further, and again in common with our approach to a reasonable percentage of revenue rate, there are two other unavoidable issues with which we must contend – procedural fairness and the power to vary a scheme. Neither party proposed using Dr Eisenach’s material in this way. Moreover, while on one view, the exercise involves nothing more than a variation to PPCA’s proposed scheme rates, on another it involves a different methodology and thus a different scheme. The tentative approach we adopted above in respect of the percentage of revenue rate thus also applies to the per stream rate (and method of arriving at that rate) which presently seems reasonable in the circumstances.

311    We know that Dr Eisenach’s purpose in these calculations was to identify the total equivalent percentage of revenue rate that would be paid to PPCA if PPCA’s per stream rates were adopted and to compare those amounts to the total amounts that would be paid to APRA. In passing, we note this seems to involve an implicit acknowledgment that the fees paid to APRA for the musical works right are relevant to the value of the right to use sound recordings in simulcasts, which is the foundation for our provisional conclusions about the percentage of revenue rates that should be paid for that right. For present purposes, however, and assuming that Dr Eisenach achieved his purpose, what is relevant is that the fees payable represent the total fees for both the simulcast right and the broadcasting right to use sound recordings.

312    We consider it likely, but are not certain, that Dr Eisenach has used the 0.4% broadcasting rate across the board in these calculations rather than attempting to use CRA’s internal breakdown of the percentage in fact paid by individual radio stations. If this is so the 0.4% can be deducted for each calculation to leave the equivalent percentage of revenue rate payable on PPCA’s per stream rates.

313    Focusing on the years 2013 to 2016 Dr Eisenach’s total percentage of revenue rates for PPCA are 1.5%, 1.8% and 2.0%. Deducting 0.4% from each would give equivalent percentage of revenue rates for the simulcast right of 1.1%, 1.4% and 1.6% for 2013/2014, 2014/2015 and 2015/2016 respectively. These percentages may be compared with those in the table at para [255] above which, in the @15% simulcast column, range from 0.0081% to 0.564% depending on music use percentage. This lack of equivalence confirms our view that the PPCA per stream rates are too high, being about twice as much as the highest percentage of revenue rate that would be paid by any radio station (one with more than 80% music use).

314    One of the advantages of a per stream rate is that there is a single rate with payments determined by actual use rather than gross proxies such as average music use percentage. Because CRA did not engage meaningfully with PPCA’s per stream rate case (as PPCA did not engage meaningfully with CRA’s percentage of revenue case) the available material requires use of the range of percentage of revenue rates which depend on music use percentages. There are many ways in which a reasonable percentage of revenue rate can be selected from that range. While we have not yet heard from the parties about any of them, obvious candidates are an overall average or the rate applicable to the average of music use which we understand to be 61%.

315    These examples of possible approaches would yield the following results:

(1)    An overall average by adding all of the rates in the simulcasting @15% column and dividing by 12:

0.296%

(2)    Overall average music use percentage of all radio stations (61%):

0.363%

316    For ease of calculation, assume the equivalent percentage of revenue rate adopted is a flat rate of 0.3%. In other words, on this assumption, the per stream rates should result in a radio station paying around 0.3% of revenue for the simulcast right. PPCA’s 2013/2014 rate is $0.0018. Dr Eisenach says this rate would equate to radio stations paying 1.5% of total revenue for both the broadcast and the simulcast right or, deducting 0.4% for the broadcast right, 1.1% for the simulcast right. If 0.3% is used instead the relevant per stream rate for that year would be roughly one third of PPCA’s proposed rate (or, as calculated, $0.00049 or 0.049 cents per stream).

317    Accordingly, if Dr Eisenach’s calculations are correct it would be possible to adjust the PPCA per stream rates downwards to achieve a per stream payment system we would consider reasonable at least at this stage.

318    We are not confident, however, that Dr Eisenach’s calculations are correct. Apart from the uncertainty as to whether he has used the 0.4% to represent the broadcast rate, he explained that he used actual data for the years 2010-2012 and then projected forward using a linear trend, as we understand it, for revenues and simulcasting audiences. As CRA pointed out:

The projected numbers start at $1.271 billion for FY14. Mr Samuel noted that the total FY12 revenue for the radio sector in Dr Eisenach's calculation was $1.12 billion. Mr Samuel commented that he did not know how Dr Eisenach produced the numbers in Exhibit 94, but raised the issue that those figures were not correctly based on the actual revenue of commercial radio stations.

319    Using the actual revenues from 2010 to 2012 and the two year compound growth for the gross earnings of the radio industry adopted by ACMA, CRA calculated that the revenues for 2013/2014 would be about 20% less than Dr Eisenach calculated. Ultimately, however, the problem, as CRA noted, is that Dr Eisenach did not present the full basis for the calculations in Exhibit 94. CRA estimated that the true equivalent percentage of revenue rate for the 2013/2014 year would be 1.75%. If that figure is used as the starting point (from which 0.4% must be deducted to give 1.35%, as opposed to 1.1% for the simulcast right), the equivalent per stream rate for 0.3% would be $0.0004 or 0.04 cents.

320    While these calculations are rough and ready (and we must give the parties a further opportunity to deal with the issues which we are raising), they yield results that are generally consistent with the views we have expressed. In particular, they support the view we have that PPCA’s rates are too high because they are based on an inappropriate starting point. They also support the view that CRA’s percentage of revenue rates for the right to use sound recordings in simulcasts of between 0.0341% and 0.065% are too low because they also are based on an inappropriate starting point.

321    For these reasons, and subject to hearing from the parties further, we are currently of the view that it would be reasonable to include in the scheme a per stream rate as PPCA proposes, the rate selected being calculated to be less for the initial year than an appropriate equivalent percentage of revenue rate to provide an incentive to radio stations to use the per stream payment system. In common with our current conclusion about the percentage of revenue payment system, we do not consider it necessary to resolve whether this would be within the power of variation and, as part of any further hearing, invite the parties to submit a scheme which assesses an appropriate per stream payment rate on this or some similar basis.

5.11    Minimum fees

322    PPCA proposes minimum fees, recoupable against the licence fee payable, as follows:

323    We do not consider the amounts proposed, particularly for metropolitan stations or regional stations with licence area more than 2% of Australian population, represent a genuine minimum fee. Instead, they represent a form of initial up-front payment of the licence fee itself. In our view, the proper purpose of a minimum fee is to ensure that PPCA’s administrative costs are covered. This is a different purpose from obtaining an up-front payment of part of the licence fee. While we do not accept Mr Samuel’s view that there should be no minimum fee at all (because it is a legitimate object to ensure that the costs of administering the scheme are covered), we are persuaded that the fees PPCA proposes are not reasonable in the circumstances.

324    As matters presently stand we cannot see any reason to require payment of a minimum fee greater than $1200 for any radio station irrespective of music use percentage, which should be able to be recouped against the licence fee payable.

5.12    PPCA’s concession

325    As part of its scheme PPCA proposed a concession so that any regional radio station with a licence area of less than 2% of the Australian population could opt out of the per stream payment system, and all associated reporting requirements, if it imposed a cap on the number of simultaneous internet connections available of 100. For such stations a flat fee of $5000 per year would apply.

326    If a per stream payment system alone were to be imposed we would consider PPCA’s concession to be reasonable and appropriate. However, as we propose that every radio station can choose whether to pay on a per stream or percentage of revenue basis we can see no justification for this concession.

5.13    Record keeping requirements

327    Consistent with the conclusions above in respect of the technical and practical issues to which a per stream payment system gives rise, we see nothing unreasonable in PPCA’s proposed record-keeping requirements for a per stream approach.

328    Insofar as the percentage of revenue payment system is concerned, the issue is affected by our views about the appropriate form of the licence which we deal with below.

5.14    The form of the licence

329    We accept PPCA’s submission that the relevant licensee is not CRA but each individual radio station.

330    Consistent with this approach, we make the following comments on the form of licence proposed:

(1)    CRA is not a relevant party to any licence granted under the scheme. Accordingly, the principal form of agreement proposed by CRA is inappropriate.

(2)    In respect of the alternative version of licence which CRA proposed, there should be no requirement that any radio station be a member of CRA reflected in the licence.

(3)    There should be no requirement that any radio station appoint CRA as its agent.

(4)    There should be no requirement for a CRA Collection Agreement between CRA and radio stations.

(5)    There should be no Gross Licence Fee. Instead each radio station should be required to pay the fee it owes to PPCA on the basis of a percentage revenue or per stream rate as decided by the radio station.

(6)    If there are confidentiality concerns about radio stations disclosing gross revenues to PPCA the licence can provide for the required information to be provided to an independent auditor (not being CRA).

(7)    Subject to these matters, there should be nothing in the licence which prevents a radio station from appointing CRA as its agent, not for the purpose of being responsible for a collection agreement based on a Gross Licence Fee, but to assist the radio station to comply with the licence.

6.    INTERIM CONCLUSIONS

331    It is convenient to express our interim conclusions in a propositional form. Our conclusions about the kind of scheme we consider would be reasonable in the circumstances are necessarily provisional because, as explained above, we consider the parties have a right to be heard about the scheme we have in mind and there should be an opportunity, if either party is so minded, for the referral of a further scheme to the Tribunal so that any issue about the power of the Tribunal to vary the schemes which have to date been referred to it need not be resolved.

332    Our interim conclusions are as follows:

(1)    None of the schemes which have been referred to us are reasonable in the circumstances.

(2)    One reason none of the schemes are reasonable is that they require payment to be made on either a per stream basis or a percentage of revenue basis and do not give radio stations the choice of which payment system they prefer. In circumstances where, first, there is a long history of payment for use of sound recordings on a percentage of revenue basis and such payments will continue under the PPCA-CRA Broadcast Agreement but also, second, there are material advantages to all parties to payment on a per stream basis, a reasonable scheme in the circumstances is one which provides a choice of payment method. The choice should be that of the radio station alone and should not operate on a “greater of” basis which would encourage the kind of gamesmanship which PPCA was concerned about for payments on a percentage of revenue basis.

(3)    Another reason all of the schemes referred to us were not reasonable in the circumstances was CRA’s schemes proposed a value for the simulcast right which was too low and PPCA’s scheme proposed a value for that right which was too high.

(4)    The polarisation of the parties about the appropriate kind of payment system and the value of the right for which payment is to be made extended to all of their forensic decisions. As such, and despite a request by the Tribunal, CRA’s evidence did not engage meaningfully with PPCA’s case (so that CRA did not suggest any per stream rate that might represent the proper value of the simulcast right) and PPCA’s evidence did not engage meaningfully with CRA’s case (so that PPCA did not suggest any percentage of revenue rate that might represent the proper value of the simulcast right). Instead, in a forensic version of mutually assured nullification, PPCA contended that there was no rational basis upon which a reasonable percentage of revenue rate could be assessed and CRA contended that there was no rational basis upon which a reasonable per stream rate could be assessed. Neither seemed concerned that if they were both right, the result would be that the Tribunal could do nothing other than reject all schemes resulting in substantial wasted time, cost and effort. We have tried to avoid that result but the consequences of so doing gives rise to issues of procedural fairness and the Tribunal’s power to vary a scheme.

(5)    As a result of the unhelpful position of the parties, in order to determine a scheme which would be reasonable in the circumstances, we have been forced to consider different methods from those proposed in the schemes the parties devised. While the result of this consideration is a scheme which permits payment on a per stream rate (as PPCA proposed) or a percentage of revenue rate (as CRA proposed), albeit the per stream rate being lower than PPCA sought and the percentage of revenue rate being higher than CRA sought, the parties must, in fairness, be given an opportunity to be heard in this regard and, if they see fit, an opportunity to address any issue about the Tribunal’s power to confirm such a scheme, if necessary, by referral of another scheme to the Tribunal.

(6)    The basic elements of a percentage of revenue payment which we currently consider would be reasonable in the circumstances involve these steps:

(a)    acceptance that the sound recording right has at least the same value as the musical works right;

(b)    identification of the current average simulcast audience (10.7%) to which must be added a percentage to reflect the unique benefits of simulcasting and the advantages it provides to radio stations; and

(c)    applying that percentage to the rates for musical works in the APRA-CRA Agreement to derive percentage of revenue rates for simulcasting.

(7)    The basic elements of a per stream payment which we currently consider would be reasonable in the circumstances involve these steps:

(a)    acceptance that the per stream rates in the NAB-SX Agreement and the various webcasting agreements are inapplicable and too high for simulcasting by radio stations in Australia;

(b)    acceptance that the per stream rate should be generally equivalent to the percentage of revenue rate we consider reasonable;

(c)    provision for some incentive to radio stations to elect to pay on a per stream basis by giving an initial discount to the per stream rate which ceases after, say, one year, when the per stream rate should increase to be equivalent to the percentage of revenue rate; and

(d)    acceptance that PPCA’s reporting requirements are basically appropriate.

(8)    Licences under the scheme should be between PPCA and each radio station. There is no requirement for a head agreement between PPCA and CRA. Moreover, the licence should not require the radio station to be a member of CRA. Nor should the licence provide for CRA to collect and distribute payments based on a gross licence fee concept. Rather, the obligation to pay should be that of each individual radio station. Concerns about confidentiality should be resolved, if need be, by interposing an independent auditor. The licence can provide for the radio station to appoint an agent for compliance purposes but should not require any radio station to do so.

333    We propose to adjourn the matter for a period of not less than 21 days so the parties have an opportunity to consider these reasons and can inform us of the directions they consider appropriate at a directions hearing thereafter.

Associate:

Dated:    28 August 2015